Developed Decision Tree Research Articles

Soil Suitability Evaluation Using Remotely Sensed Data and GIS: A Case Study from Kumaon Himalayas

Soil resource characterization of Kumaon hills of Himalayas was carried out through soil survey and established soil–physiography relationship after identification and delineation of six physiographic units in Khulgad watershed of Almora district, Uttarakhand. The satellite image characteristics identified through the visual/digital interpretation of IRS-1D LISS-III fused with PAN data on 1:12,500 scale were utilized in the present study. Based on the characteristics of soils developed on different physiographic units, the soils were evaluated for land capability classification and suitability for major crops in the region for sustainable development of the study area. GIS technique was used to develop decision trees employing various soil–site parameters for assessing crop suitability. The results revealed that deep to moderately deep soils with high clay content on the stable terraces and valley lands mostly appearing as dark blue to blackish blue in satellite imagery grouped under land capability classes II and III were interpreted as suitable for wheat, rice, finger millet and mustard crops. Moderately deep to deep soils occurring on moderately steep hill slopes and upper terraces were suitable for potato, mustard and finger millets. Soils on degraded hills with terraced tops having forest in patches (showed mix tone of grey and pink in imagery) categorized into class IV were found to be marginally suitable for wheat, mustard and potato with constraints of soil and topography, while soils on upper hill slopes, degraded hill tops/slopes with thin forest/grazing land and rocky ridges with dark to light blue, white and greyish mixed tone and dark blue with pale pink mixed tone in the imagery placed under land capability classes VI to VII were presently not suitable for crops cultivation and moderately suitable for grassland pasture or silvipasture purposes. The extent of suitability of soils for various crops has also been highlighted using GIS software and the maps were prepared. GIS is found to be helpful in processing spatial and non-spatial remote sensing data and conventional data of highly folded and faulted chain of Kumaon Himalayas to evaluate soil–site parameters for assessing crop suitability for optimum utilization of land.

Cost-effectiveness of Abdominal Ultrasound Versus Magnetic Resonance Imaging for Pancreatic Cancer Screening in Familial High-Risk Individuals in Japan.

The aim of the study was to assess cost-effectiveness of abdominal ultrasound, magnetic resonance imaging, endoscopic ultrasound, computed tomography, positron emission tomography, and no screening for pancreatic cancer screening in familial high-risk individuals (HRIs). We developed decision trees with Markov models for a hypothetical cohort of familial HRIs at the age of 50 year using a healthcare sector perspective and a lifetime horizon. Main outcomes were costs, quality-adjusted life-years (QALYs), and incremental cost-effectiveness ratios. In a base-case analysis, abdominal ultrasound was the most cost-effective (US $11,035, 17.4875 QALYs). Magnetic resonance imaging yielded the best benefits. Cost-effectiveness was sensitive to the incidence of pancreatic cancer. Endoscopic ultrasound was more cost-effective than abdominal ultrasound when the incidence of pancreatic cancer was greater than 0.008 and under 0.016. Magnetic resonance imaging was more cost-effective than endoscopic ultrasound when the incidence of pancreatic cancer was greater than 0.016. Probabilistic sensitivity analysis using Monte-Carlo simulation for 10,000 trials demonstrated that abdominal ultrasound was cost-effective 76% of the time at a willingness-to-pay threshold of US $50,000/QALY gained. Abdominal ultrasound is the most cost-effective and recommended for pancreatic cancer screening in familial HRIs in Japan. Evaluating the risk of pancreatic cancer among familial HRIs as a target for screening is significant.

Breathing new life into postmortem analysis: Identification of key failure modes for conventional petroleum segments offshore New Zealand

Technically rigorous, systematic, and consistent postmortem analysis of dry holes is necessary to improve future exploration performance. Here, we identify the key failure modes for three segments in conventional petroleum prospects using a previously developed decision tree. The failed segments were tested by wells Whio-1 and Kanuka-1 in the Taranaki Basin, offshore New Zealand. Using comprehensive geological, geophysical and geochemical data from the wells, we determined that the segments failed due to lack of petroleum migration (segment M2A in well Whio-1 and segment Mt. Messenger in well Kanuka-1) and due to unfavorable timing of petroleum migration (segment Mangahewa in well Whio-1). These failure modes are broadly aligned with those determined by the operators of these wells. However, the use of decision tree results in more consistent and less vague definitions of failure modes.

Data Mining Generating Decision Trees to Alert System Against Death and Losses in Egg Production

Climatic changes and high temperatures have been affecting animal production and the well-being of laying birds, with heat stress and high mortality rates, generating economic losses. Legacy databases can contain information to help model thermal comfort at climatic extremes. They can enable decision trees to be created through the use of data mining to prevent mortality and production losses. Thus, the objective of this study is to seek to develop decision trees, for application as an alert system, for the incidence of caloric stress in the production of layers. We used a database of three aviaries located in the city of Bastos-SP, collected in 2013. The data were organized in Excel® spreadsheets, and processed with the Weka® software with the J48 (C4.5) algorithm for mining of the data. The technique allowed the construction of decision trees that in the chosen sheds were classified with respectively 99.73%, 99.61%, and 98.71% of correct answers and with Kappa indexes equal to 0.9958, 0.9907 and 0.9663, which indicate that the three classifiers built are excellent. Thus, the proposed system, with the decision trees built, can serve as a basis for the construction of an alert system to be applied to the three warehouses simultaneously.

A machine learning approach to discriminate MR1 binders: The importance of the phenol and carbonyl fragments

AimsIn this study, we attempted to discriminate between MR1 binders and non-binders using machine learning (ML) approach and emphasized the important descriptors. Background: The major histocompatibility complex (MHC) class I-related molecule, MR1, is a component of the Immune system and interacts with T cell receptor (TCR) to modulate the immune response against various antigens. MR1 has raised many interests in recent years due to the potential of presenting a broader range of small molecules. MR1 has a small ligand-binding pocket interacting with agonistic or antagonistic ligands to stimulate or inhibit the immune response, respectively. Objective: There are limited studies on designing small molecules for the MR1 binding site, and the available raw data for MR1 binders is insufficient to exploit them for prioritizing chemicals. Therefore, the objective of this study was to provide validated and precise outcomes to expand the knowledge of critical structural features of MR1 binders. Method: We developed QSAR classifier models using Decision Tree (DT), Artificial Neural Network (ANN), Random Forest (RF), Extra Tree (ET), Linear Support Vector Machine (LSVM), Logistic Regression (LR), Naïve Bayesian classification (NB), and K-nearest-neighbors (KN). Result: The total accuracies for the best Machine Learning (ML) models were over 85%. The developed Decision Tree (DT) using suggested descriptors (fr_C_O_noCOO, fr_phenol, PEOE_VSA2) was able to classify the binders and non-binders with the accuracy of 85% for the train set and 100% for the test set. However, the 100% accuracy might be achieved by chance (due to simple random split of train/test set). DT models are easily interpretable. Therefore, a set of simple association rules was provided based on the provided DT model. Moreover, a LR equation was provided. Conclusion: The developed DT and LR models and rules could be used directly for ligand optimization, virtual screening, or re-scoring structure-based virtual screening results after consideration of the domain applicability. In general, the most important descriptors were found to be fr_C_O_noCOO, fr_phenol, PEOE_VSA2 and to lesser extents, NumHDonors and VSA_Estate8 that were consistent with available crystallographic structures.

Identifier of Regional Food Presence (IRFP): A New Perspective to Evaluate Sustainable Menus

The menu offered in restaurants must meet different aspects of quality. Cultural elements are related to their acceptance and can contribute to the preservation of habits, sustainable agricultural systems, and the maintenance of biodiversity and sustainability, among other factors. In this context, this research proposes an instrument for classifying menus regarding the presence/absence of regional foods called the identifier of regional foods presence (IRFP) as a new perspective to evaluate sustainable menus. For this, lists of regional preparations and ingredients were prepared for each Brazilian region. Sequentially, we submitted the dishes/ingredients to a developed decision tree for the classification of foods into regional or national foods. The score, based on the presence/absence of regional foods, considered the components of a menu, with zero attributed to a lack of regional ingredients/dishes. For national dishes/ingredients, researchers attributed a minimum score equal to ten. One regional food gave a score of50 to the menu, and with more than one regional food, a daily menu scored 100. The final menu evaluation was based on the mean scores of the menus in each restaurant. Scores between 0–49.9 were considered inadequate; 50–74.9, adequate; and excellent between 75–100. The IRFP was applied to 111 menus with data collected from all the offered dishes. In total, the study evaluated data from 774 recipes from the menus of 37 restaurants located in the five Brazilian regions with a similar operating system. ANOVA was used to verify if there was a statistical difference between the mean score of each Brazilian region (p < 0.05). The average score obtained by the IRFP in menus from Brazilian community restaurants was 80.3 ± 30.9 (excellent), showing a significant difference between the Northeast and Southeast Regions, with a more significant presence of regional foods in the Northeast (87.7 ± 28.7). The use of the IRFP in menus was shown to be easy in its application, contributing to a stimulation of the use of regional items and, consequently, to the direct and indirect benefits generated for the food system and the local population.

Contrast-enhanced ultrasound features of malignant focal liver masses in dogs

A total of 185 cases (150 retrospectively and 35 prospectively) of malignant liver masses were collected. In the retrospectively collected cases hyperenhancement during wash-in was the most common feature in HCCs but there was a high percentage of cases showing no enhancement or hypo/isoenhancement. ICCs displayed a large variety of contrast enhancement patterns and, although statically significant differences between ICCs and HCCs were evident, no clear distinction between these two pathologies was possible based only on their CEUS appearance. Sarcomas displayed all the possible degrees of wash-in enhancement with non-enhancing being the most common appearance. Metastases displayed all the possible contrast-enhancement patterns, with the most common being hyperenhancement in the wash-in phase followed by hypoenhancement in the wash-out phase. A decision tree was developed based on the features of the retrospectively selected cases. Based on the developed decision tree 27/35 prospectively collected cases were correctly classified. Even if some significant differences among groups were evident, all the histotypes displayed all the possible patterns of contrast enhancement, and, therefore, the differentiation of liver masses in dogs based only on their CEUS features is not feasible and, therefore, cytology or histopathology is required.

Developing Decision Tree based Models in Combination with Filter Feature Selection Methods for Direct Marketing

Direct Marketing is a form of advertising strategies which aims to communicate directly with the most potential customers for a certain product using the most appropriate communication channel. Banks are spending a huge amount of money on their marketing campaigns, so they are increasingly interested in this topic in order to maximize the efficiency of their campaigns, especially with the existence of high competition in the market. All marketing campaigns are highly dependent on the huge amount of available data about customers. Thus special Data Mining techniques are needed in order to analyze these data, predict campaigns efficiency and give decision makers indications regarding the main marketing features affecting the marketing success. This paper focuses on four popular and common Decision Tree (DT) algorithms: SimpleCart, C4.5, RepTree and Random Tree. DT is chosen because the generated models are in the form of IF-THEN rules which are easy to understand by decision makers with poor technical background in banks and other financial institutions. Data was taken from a Portuguese bank direct marketing campaign. A filter-based Feature selection is applied in the study to improve the performance of the classification. Results show that SimpleCart has the best results in predicting the campaigns success. Another interesting finding that the five most significant features influencing the direct marketing campaign success to be focused on by decision makers are: Call duration, offered interest rate, number of employees making the contacts, customer confidence and changes in the prices levels.

Decision Trees for Predicting the Physiological Responses of Rabbits.

Simple SummaryThe primary aim of this paper is to develop decision trees to predict rabbits’ physiological responses, such as the respiratory rate or ear temperature, based on environmental variables (dry bulb temperature and relative humidity). The decision tree for ear temperature exhibited better statistical indices, indicating the benefits of using the ear temperature as an indicator of thermal stress. Our findings confirm that the resulting decision trees are powerful classifiers, and the results can be easily understood. Hence, the proposed decisions trees can aid in investigating the influence of environmental conditions on physiological responses and, consequently, the rabbits’ welfare. These results can be used in practical situations and can be obtained in real time to support rabbit breeders in decision-making to improve environmental conditions for rabbits. The thermal environment inside a rabbit house affects the physiological responses and consequently the production of the animals. Thus, models are needed to assist rabbit producers in decision-making to maintain the production environment within the zone of thermoneutrality for the animals. The aim of this paper is to develop decision trees to predict the physiological responses of rabbits based on environmental variables. The experiment was performed in a rabbit house with 26 rabbits at eight weeks of age. The experimental database is composed of 546 observed data points. Sixty decision tree models for the prediction of respiratory rate (RR, mov.min−1) and ear temperature (ET, °C) of rabbits exposed to different combinations of dry bulb temperature (tdb, °C) and relative humidity (RH, %) were developed. The ET model exhibited better statistical indices than the RR model. The developed decision trees can be used in practical situations to provide a rapid evaluation of rabbit welfare conditions based on environmental variables and physiological responses. This information can be obtained in real time and may help rabbit breeders in decision-making to provide satisfactory environmental conditions for rabbits.

Design and implementation of green chemistry approaches into pharmaceutical analysys of benzydamine dosage formes

Aim. The development of the pharmaceutical industry of Ukraine and the world has led to an increase in the need for the use of hazardous and toxic chemicals and solvents, which affects the safety of the environment and directly employees of pharmaceutical companies.Therefore, one of the solutions to this problem is the implementation of “green chemistry” approaches in pharmaceutical quality control laboratories.Materials and methods. Chromatographic separation methods are used for the qualitative and quantitative analysis of raw materials and finished dosage forms, the determination of substances that are formed during the degradation of active substances and allow rapid analysis of complex mixtures.Results. For the implementation of principles in the laboratory of pharmaceutical companies, it is necessary to evaluate the possibility of using rapid quality control methods such as gas chromatography, ultra-high performance liquid chromatography, and absorption spectrophotometry in the ultraviolet and visible regions.Conclusions. Approaches to greening of analytical procedures used in quality control of pharmaceuticals have been studied. Ways of implementation of modern approaches of methods of green chemistry to chromatographic methods are offered. On the basis of the developed decision tree the design of development and greening of the methods of quality control of benzidamine dosage forms is proposed

MAIA (Medical Artificial Intelligence Assistant) as interface for a new cancer healthcare integrative platform.

25 Background: Cancer healthcare systems are an example of inequity and waste in low and middle income countries. Access to high quality cancer pathways focused in early diagnosis, molecular biology, proper staging and evidence based treatments are scarce and the patient`s care experience is dramatic and difficult in a majority of cases. There are no integrative healthcare models based on new technologies that improve outcomes and make more comfortable and expeditious all the patient and physician´s journey in cancer. Methods: Our team developed and trained a talkbot called MAIA (Medical Artificial Intelligence Assistant) using an algorithmic translation of medical language focused in the state or art for non small cell lung cancer. Our clinical team developed decision trees in diagnosis, staging, medical and surgical treatment and molecular biology that were incorporated in a virtual platform and then integrated onto a narrow artificial intelligence bot brain using neural networks with the proposal of generate clinical support to the physician and create a standard text using the verbal information captured in the oncological consultation and integrated images (reports) through a image edition software and then create a unique medical record without using computers by the physician. MAIA also can create medical treatment choices in first line of treatment and create alerts and alarms through an own app (MAIA Hip). Results: Our proof of concept was released in video at this link https://drive.google.com/file/d/12YtiOkhfEmIsL2bFp9T3QyfHHWxBvvKU/view?ts=5ceec096 Due to our decision trees size we can´t upload them, but are available for presentation. Conclusions: A talkbot trained as a narrow artificial intelligence interface for an integrative cancer healthcare platform (HIP) is possible through the clinical and engineer integration of languages using a neural network method and other software tools. MAIA is for now a patient and physician experience improvement, but the real impact will be in the data standarization and acquisition for advanced analytics. The final scope of MAIA HIP will be a blockchain for cancer in low and middle income countries.

Development of decision trees to discriminate HDAC8 inhibitors and non-inhibitors using recursive partitioning

Histone deacetylase 8 (HDAC8) is involved in malignancy. Overexpression of HDAC8 is correlated with various cancers. Design of selective HDAC8 inhibitors is always a challenging task to the chemistry audiences. In this communication, a diverse set comprising large number of compounds are subjected to recursive partitioning (RP) analysis to develop decision trees to discriminate compounds into HDAC8 inhibitors (active) and non-inhibitors (inactive). Acquiring knowledge about the essential structural and physicochemical parameters can be useful in designing potential and selective HDAC8 inhibitors. Moreover, this work validates our previous results observed in Bayesian modelling study of this dataset. This comparative learning will surely enrich drug discovery aspects related to HDAC8 inhibitors.Communicated by Ramaswamy H. Sarma

Cost-effectiveness of Helicobacter pylori test and eradication versus upper gastrointestinal series versus endoscopy for gastric cancer mortality and outcomes in high prevalence countries

Background: The latest version of Japanese guidelines for effective secondary prevention of gastric cancer recommend upper gastrointestinal series (UGI) and endoscopy in adults 50 years of age and older. A Helicobacter pylori antibody test and eradication (H. pylori screening) reduces gastric cancer risk.Objective: This study aimed to evaluate the cost-effectiveness of H. pylori screening, compared to UGI and endoscopy in high prevalence countries.Methods: We developed decision trees with Markov models using a healthcare payer perspective and a lifetime horizon. Targeted populations were hypothetical cohorts of asymptomatic individuals at the age of 50, 60, 70 and 80 years. We calculated per-person costs and effectiveness with discounting at a fixed annual rate of 3% and compared incremental cost-effectiveness ratios.Results: H. pylori screening was cost-saving and more cost-effective for individuals at the age of 50, 60, 70, and 80 than UGI and endoscopy. One-way and multiway sensitivity analyses showed the robustness of the cost-effectiveness results. Probabilistic sensitivity analyses using Monte-Carlo simulation for 10,000 trials demonstrated that H. pylori screening was cost-effective 100% of the time at a willingness-to-pay level of US$50,000/QALY gained.Conclusions: H. pylori screening for the adults 50 years of age and older could be cost-effective compared to UGI and endoscopy in high prevalence countries. The main reasons for the superiority of H. pylori screening are that an H. pylori antibody test has a higher sensitivity and specificity than UGI and endoscopy and the benefits to reduce gastric cancer incidence and mortality.

Classification of air traffic control scenarios using decision trees: insights from a field study in terminal approach radar environment

Air traffic controllers are responsible for the safe, expeditious and orderly flow of the air traffic. Their training relies heavily on the use of simulators that can represent various normal and emergency situations. Accurate classification of air traffic scenarios can provide assistance towards a better understanding of how controllers respond to the complexity of a traffic scenario. To this end, we conducted a field study using qualified air traffic controllers, who participated in simulator sessions of terminal radar approach control in a variety of scenarios. The aim of the study was twofold, firstly to explore how decision trees and classification rules can be used for realistic classification of air traffic scenarios and secondly to explore which factors reflect better operational complexity. We applied machine learning methods to the data and developed decision trees and classification rules for these scenarios. Results indicated that decision trees and classification rules are useful tools in accurately categorizing scenarios and that complexity requires a larger set of predictors beyond simple aircraft counts. The derived decision trees and classification rules performed well in prediction, stability and interpretability. Practical benefits can be derived in the areas of operations and system design in the context of air traffic flow and capacity management systems.

Use of Germline Genetic Variability for Prediction of Chemoresistance and Prognosis of Breast Cancer Patients.

The aim of our study was to set up a panel for targeted sequencing of chemoresistance genes and the main transcription factors driving their expression and to evaluate their predictive and prognostic value in breast cancer patients. Coding and regulatory regions of 509 genes, selected from PharmGKB and Phenopedia, were sequenced using massive parallel sequencing in blood DNA from 105 breast cancer patients in the testing phase. In total, 18,245 variants were identified of which 2565 were novel variants (without rs number in dbSNP build 150) in the testing phase. Variants with major allele frequency over 0.05 were further prioritized for validation phase based on a newly developed decision tree. Using emerging in silico tools and pharmacogenomic databases for functional predictions and associations with response to cytotoxic therapy or disease-free survival of patients, 55 putative variants were identified and used for validation in 805 patients with clinical follow up using KASPTM technology. In conclusion, associations of rs2227291, rs2293194, and rs4376673 (located in ATP7A, KCNAB1, and DFFB genes, respectively) with response to neoadjuvant cytotoxic therapy and rs1801160 in DPYD with disease-free survival of patients treated with cytotoxic drugs were validated and should be further functionally characterized.

Comparison of Landscape Metrics for Three Different Level Land Cover/Land Use Maps

This research aims to investigate how different landscape metrics are affected by the enhancement of the thematic classes in land cover/land use (LC/LU) maps. For this aim, three different LC/LU maps based on three different levels of CORINE (Coordination of Information on The Environment) nomenclature were created for the selected study area using GEOBIA (Geographic Object Based Image Analysis) techniques. First, second and third level LC/LU maps of the study area have five, thirteen and twenty-seven hierarchical thematic classes, respectively. High-resolution Spot 7 images with 1.5 m spatial resolution were used as the main Earth Observation data to create LC/LU maps. Additional geospatial data from open sources (OpenStreetMap and Wikimapia) were also integrated to the classification in order to identify some of the 2nd and 3rd level LC/LU classes. Classification procedure was initially conducted for Level 3 classes in which we developed decision trees to be used in object-based classification. Afterwards, Level 3 classes were merged to create Level 2 LC/LU map and then Level 2 classes were merged to create the Level 1 LC/LU map according to CORINE nomenclature. The accuracy of Level 1, Level 2, Level 3 maps are calculated as; 93.50%, 89.00%, 85.50% respectively. At the last stage, several landscape metrics such as Number of Patch (NP), Edge Density (ED), Largest Patch Index (LPI), Euclidean Nearest Neighbor Distance (ENN), Splitting Index (SPLIT) and Aggregation Index (AI) metrics and others were calculated for different level LC/LU maps and landscape metrics values were compared to analyze the impact of changing thematic details on landscape metrics. Our results show that, increasing the thematic detail allows landscape characteristics to be defined more precisely and ensure comprehensive assessment of cause and effect relationships between classes.

Preoperative and postoperative prediction of long-term meningioma outcomes.

BackgroundMeningiomas are stratified according to tumor grade and extent of resection, often in isolation of other clinical variables. Here, we use machine learning (ML) to integrate demographic, clinical, radiographic and pathologic data to develop predictive models for meningioma outcomes.Methods and findingsWe developed a comprehensive database containing information from 235 patients who underwent surgery for 257 meningiomas at a single institution from 1990 to 2015. The median follow-up was 4.3 years, and resection specimens were re-evaluated according to current diagnostic criteria, revealing 128 WHO grade I, 104 grade II and 25 grade III meningiomas. A series of ML algorithms were trained and tuned by nested resampling to create models based on preoperative features, conventional postoperative features, or both. We compared different algorithms’ accuracy as well as the unique insights they offered into the data. Machine learning models restricted to preoperative information, such as patient demographics and radiographic features, had similar accuracy for predicting local failure (AUC = 0.74) or overall survival (AUC = 0.68) as models based on meningioma grade and extent of resection (AUC = 0.73 and AUC = 0.72, respectively). Integrated models incorporating all available demographic, clinical, radiographic and pathologic data provided the most accurate estimates (AUC = 0.78 and AUC = 0.74, respectively). From these models, we developed decision trees and nomograms to estimate the risks of local failure or overall survival for meningioma patients.ConclusionsClinical information has been historically underutilized in the prediction of meningioma outcomes. Predictive models trained on preoperative clinical data perform comparably to conventional models trained on meningioma grade and extent of resection. Combination of all available information can help stratify meningioma patients more accurately.

Restoring historical forest conditions in a diverse inland Pacific Northwest landscape

AbstractA major goal of managers in fire‐prone forests is restoring historical structure and composition to promote resilience to future drought and disturbance. To accomplish this goal, managers require information about reference conditions in different forest types, as well as tools to determine which individual trees to retain or remove to approximate those reference conditions. We used dendroecological reconstructions and General Land Office records to quantify historical forest structure and composition within a 13,600 ha study area in eastern Oregon where the USDA Forest Service is planning restoration treatments. Our analysis demonstrates that all forest types present in the study area, ranging from dry ponderosa pine‐dominated forests to moist mixed conifer forests, are considerably denser (273–316% increase) and have much higher basal area (60–176% increase) today than at the end of the 19th century. Historically, both dry pine and mixed conifer forest types were dominated by shade‐intolerant species. Today, shade‐tolerant tree cover has increased in dry pine stands, while mixed conifer stands are now dominated by shade‐tolerant species. Federal managers in eastern Oregon are currently required to retain all live trees >53 cm diameter at breast height in the course of forest management activities because this size class is assumed to be under‐represented on the landscape relative to historical conditions. However, we found the same or greater number of live trees >53 cm today than in the late 19th century. Restoring historical conditions usually involves removing shade‐tolerant trees that established since Euro‐American management significantly altered natural disturbance regimes. We evaluated a wide range of tree morphological and environmental variables that could potentially predict the age of grand fir and Douglas‐fir, the most abundant shade‐tolerant species found within the study area. We describe several morphological characteristics that are diagnostic of tree age and developed decision trees that predict the approximate age of trees using morphological characteristics that are easy to measure in the field such as height to live foliage or height to dead branches. Information about structural and compositional change over time combined with tree‐age prediction tools provides a flexible framework for restoring historical conditions and meeting other resource management objectives.

Risk-based monitoring of chemical substances in food: Prioritization by decision trees

Monitoring programs for food safety hazards are increasingly set up on a risk basis. In order to include the most relevant chemical substances for food products within such a program, a prioritization is needed. Within this research, decision trees were developed to rank chemical substances that may occur in food products. The decision trees focus on red meat, but can also be applied for other animal products. Three decision trees were developed, one for each of the following groups of chemical substances: prohibited substances, environmental contaminants and authorized substances. The decision trees focused on human health risks including the occurrence of a chemical substance and the human health effects. Substances are subsequently classified into high, medium and low priority for monitoring. The decision trees were tested with a selection of chemical substances for each of the three groups of chemical substances. Scientific information (such as EFSA opinions) and (national) monitoring data were used to answer the questions in the decision trees. Results showed that the developed decision trees are helpful in classifying chemical substances and as such can be used as a tool to set up risk based monitoring plans. It is recommended to establish a working group containing a broad range of experts to prioritize the substances and to re-evaluate the list of prioritized substances frequently using the most current (monitoring) information.

Identifying Predictable Patterns in the Unconventional Overtaking Decisions of PTW for Cooperative ITS

In the modern challenging urban environment, powered two wheelers (PTW) are selected for their everyday commuting. The way motorcycles and scooters travel through traffic is systematically emphasized in the recent relevant literature as being complex, especially in relation to overtaking. In this paper, meta-optimized Decision Trees, a special case of machine learning (ML) models, are developed in order to model the unconventional overtaking patterns of PTW drivers. Based on detailed naturalistic trajectory data collected using video footage from unmanned aerial vehicles (UAV) in a three-lane arterial in Athens, Greece, two different models of PTW driving behavior are developed. The first model addresses the decision of the PTW driver to overtake or not the preceding vehicle. The second model focuses on PTW driver's intention to overtake or undertake (pass from the right) it. The developed decision tree models are further analyzed in relation to the revealed significant factors during overtaking. Following, the applicability of the developed algorithms in the context of intelligent transportation systems and connected vehicles is discussed, which reveals the importance of acquiring quality data using advanced equipment combined with advanced ML approaches for advanced modeling methods.