Machine Learning Algorithms Decision Tree Research Articles

CLASSIFICATION OF KAZAKH MUSIC GENRES USING MACHINE LEARNING TECHNIQUES

This article analysis a Kazakh Music dataset, which consists of 800 audio tracks equally distributed across 5 different genres. The purpose of this research is to classify music genres by using machine learning algorithms Decision Tree Classifier and Logistic regression. Before the classification, the given data was pre-processed, missing or irrelevant data was removed. The given dataset was analyzed using a correlation matrix and data visualization to identify patterns. To reduce the dimension of the original dataset, the PCA method was used while maintaining variance. Several key studies aimed at analyzing and developing machine learning models applied to the classification of musical genres are reviewed. Cumulative explained variance was also plotted, which showed the maximum proportion (90%) of discrete values ​​generated from multiple individual samples taken along the Gaussian curve. A comparison of the decision tree model to a logistic regression showed that for f1 Score Logistic regression produced the best result for classical music - 82%, Decision tree classification - 75%. For other genres, the harmonic mean between precision and recall for the logistic regression model is equal to zero, which means that this model completely fails to classify the genres Zazz, Kazakh Rock, Kazakh hip hop, Kazakh pop music. Using the Decision tree classifier algorithm, the Zazz and Kazakh pop music genres were not recognized, but Kazakh Rock with an accuracy and completeness of 33%. Overall, the proposed model achieves an accuracy of 60% for the Decision Tree Classifier and 70% for the Logistic regression model on the training and validation sets. For uniform classification, the data were balanced and assessed using the cross-validation method. The approach used in this study may be useful in classifying different music genres based on audio data without relying on human listening.

High-performance grating-like SERS substrate based on machine learning for ultrasensitive detection of Zexie-Baizhu decoction

Rapid, universal and accurate identification of chemical composition changes in multi-component traditional Chinese medicine (TCM) decoction is a necessary condition for elucidating the effectiveness and mechanism of pharmacodynamic substances in TCM. In this paper, SERS technology, combined with grating-like SERS substrate and machine learning method, was used to establish an efficient and sensitive method for the detection of TCM decoction. Firstly, the grating-like substrate prepared by magnetron sputtering technology was served as a reliable SERS sensor for the identification of TCM decoction. The enhancement factor (EF) of 4-ATP probe molecules was as high as 1.90 × 107 and the limit of detection (LOD) was as low as 1 × 10−10 M. Then, SERS technology combined with support vector machine (SVM), decision tree (DT), Naive Bayes (NB) and other machine learning algorithms were used to classify and identify the three TCM decoctions, and the classification accuracy rate was as high as 97.78 %. In summary, it is expected that the proposed method combining SERS and machine learning method will have a high development in the practical application of multi-component analytes in TCM.

Analysis and Prediction of Students' Adaptation to Online Education Systems Based on Data Analysis and Decision Tree Machine Learning Algorithms

In today's digital age, the popularity and development of online education systems provide students with more flexible and convenient ways of learning. However, students' adaptation to the online education system is affected by a variety of factors, including gender, age, educational background, and field of specialisation. Through in-depth analyses and studies of these factors, the following conclusions can be drawn: gender has little influence on students' adaptation to online education, and male and female students perform similarly overall, but the proportion of male students at high adaptation levels is significantly higher than that of females. The majority of students show medium adaptability, indicating that the overall effect of online education is average. students in the age groups of 6-10, 16-20 and 26-30 years old have lower adaptability levels, and there are more low adaptability groups among students in colleges and universities. students majoring in IT are more adapted to the online education system, and students not majoring in IT have relatively poorer adaptability level. Local students are more adaptable to online education than foreign students. In areas with unstable electricity, students' adaptability is usually lower. The decision tree algorithm predictions showed good overall model accuracy, with higher prediction accuracy for students with high, low and medium levels of adaptability. The test set accuracy was 93.27%, and the precision and recall were both 93.33%, indicating excellent model predictions. In summary, by deeply analysing the influence of various factors on students' adaptation degree to online education and using the random forest algorithm to make predictions, it can provide an important reference for improving the effectiveness of online education systems and provide useful insights for personalised education.

Multiple Disease Prediction Using Machine Learning

Machine learning, which is a type of computer technology, has changed healthcare a lot. It helps doctors predict diseases better and faster. In healthcare, using machine learning algorithms decision tree (DT), logistic regression (LR), support vector machine (SVM) that can help predict lots of different diseases at the same time. This helps doctors find and treat illnesses early, which makes patients better and saves money on healthcare. This paper looks at how we can use computer programs that learn from data to predict many diseases. It talks about why this is good, what problems we might face, and where we might go next with it. We give a summary of the several machine learning models and information sources that are often employed in illness prediction. We also go over the significance of feature selection, model assessment, and combining several data modalities for improved illness prediction. We give a summary of the several machine learning models and information sources that are often employed in illness prediction. We also go over the significance of feature selection, model assessment, and combining several data modalities for improved illness prediction. The research shows that using machine learning algorithms to predict many diseases at once could really help public health. Again, we use a machine learning model to determine whether or not an individual is impacted by a few diseases. This training model trains itself to predict illness using sample data.

Creditworthiness pattern prediction and detection for GCC Islamic banks using machine learning techniques

PurposeBank creditworthiness refers to the evaluation of a bank’s ability to meet its financial obligations. It is an assessment of the bank’s financial health, stability and capacity to manage risks. This paper aims to investigate the credit rating patterns that are crucial for assessing creditworthiness of the Islamic banks, thereby evaluating the stability of their industry.Design/methodology/approachThree distinct machine learning algorithms are exploited and evaluated for the desired objective. This research initially uses the decision tree machine learning algorithm as a base learner conducting an in-depth comparison with the ensemble decision tree and Random Forest. Subsequently, the Apriori algorithm is deployed to uncover the most significant attributes impacting a bank’s credit rating. To appraise the previously elucidated models, a ten-fold cross-validation method is applied. This method involves segmenting the data sets into ten folds, with nine used for training and one for testing alternatively ten times changeable. This approach aims to mitigate any potential biases that could arise during the learning and training phases. Following this process, the accuracy is assessed and depicted in a confusion matrix as outlined in the methodology section.FindingsThe findings of this investigation reveal that the Random Forest machine learning algorithm superperforms others, achieving an impressive 90.5% accuracy in predicting credit ratings. Notably, our research sheds light on the significance of the loan-to-deposit ratio as a primary attribute affecting credit rating predictions. Moreover, this study uncovers additional pivotal banking features that intensely impact the measurements under study. This paper’s findings provide evidence that the loan-to-deposit ratio looks to be the purest bank attribute that affects credit rating prediction. In addition, deposit-to-assets ratio and profit sharing investment account ratio criteria are found to be effective in credit rating prediction and the ownership structure criterion came to be viewed as one of the essential bank attributes in credit rating prediction.Originality/valueThese findings contribute significant evidence to the understanding of attributes that strongly influence credit rating predictions within the banking sector. This study uniquely contributes by uncovering patterns that have not been previously documented in the literature, broadening our understanding in this field.

Investigation of light gas adsorption by microporous materials using data analysis and decision tree machine learning algorithm

This study aimed to extract insights into the adsorption behavior of light gases by microporous materials such as zeolites, MOFs, and activated carbons. Data reported in 22 articles published between 1974 and 2022 were analyzed using a decision tree (DT) machine learning algorithm. A comprehensive database comprising 3297 data points, elucidating the impacts of 8 input variables on adsorption capacity, was constructed. Various exploratory data analysis techniques, including histograms, bar charts, and scatter plots, were employed to discern the relationships among input variables and the performance variable. Additionally, a DT model was employed to regress the adsorbed capacity data. Furthermore, a parametric study of this model facilitated the determination of the relative importance of input variables and their partial dependence, enhancing the interpretability of the model and enabling the deduction of heuristics for high or low adsorption capacity. The exploratory data analysis revealed that pressure and molecular mass of the adsorbate were the most significant variables influencing adsorption capacity.

Predicting and analyzing the fill factor of non-fullerene organic solar cells based on material properties and interpretable machine-learning strategies

In recent years, encouraging progress in non-fullerene acceptors-based organic solar cells (NFAs-OSCs) has been made in the field of clean-energy technologies. Nevertheless, achieving a high fill factor (FF) for NFAs-OSCs is a great challenge due to the FF value can be dramatically affected by various factors. An accurate prediction (based on empirical data) of how inherent characteristics of polymer-NFA combinations (e.g., frontier molecular orbitals (FMO) and charge-carrier mobilities) affects the FF in NFAs-OSCs is therefore urgently needed for the future commercialization prospect. The presented work demonstrates an outperforming predictive model using the optimized Gradient-boosting decision tree (GBDT) machine learning (ML) algorithm to accurately predict the FF value, which is based on a dataset consisting of > 180 unique donor/NFA blends with reported FF from previously published publications and experimental physical descriptors (FMO and charge-carrier mobilities). In addition, the SHapley Additive exPlanations (SHAP) strategy is further applied to not only gain the interpretability of the GBDT model prediction but also directly map the relationship between the FF value and experimental physical descriptors by systematic interpretable explanations (e.g., global and local). The SHAP analysis qualitatively reveals that the highest occupied molecular orbital (HOMO) of the donor and the HOMO of the acceptor are the two most important features associated with FF value prediction, which is in good agreement with previous experimental facts. Furthermore, the ML strategies explore the key requirements (e.g., small HOMO offset (<0.3 eV), high hole and electron mobilities (>1.0 × 10–4 cm2 V−1 s−1), and desired μh/μe ratio (in the range of 1.8–3.3)) for enhancing the FF values of NFAs-OSCs. This study provides novel insights into the possible physical mechanisms of FF enhancement, thus facilitating research efforts (e.g., rational design of energy level tuning in donor/NFA blend) for high-performance NFAs-OSCs.

Analyzing electric vehicle battery health performance using supervised machine learning

Lithium-ion batteries having high energy and power densities, fast depleting cost, and multifaceted technological improvement lead to the first choice for electric transportation systems. A noble supervised K nearest neighbors (KNN), support vector regressor (SVR), decision tree (DT), and random forest (RF) regressors machine learning algorithm are developed with different accuracy and usefulness for the state of health estimation from direct measurable indices of voltage, current, and temperature without inherited electrochemical characteristics like voltage hysteresis, aging, degradation level, operational, and environmental effect. Based on the battery dataset from Sandia National Laboratories, the proposed technique is validated. The comparison characteristics are provided with the help of mean absolute percent error (MAPE), mean squared error (MSE), root mean squared error (RMSE), and mean absolute error (MAE) techniques. The MAPE error at 15 °C for different regression algorithms SVR, DT, KNN, and RF are 2.791E-02, 1.607E-03, 9.957E-04, and 4.323E-03 respectively. The MAE error at 25 °C for different regression algorithms SVR, DT, KNN, and RF are 5.593E-02, 2.379E-03, 2.429E-03, and 5.073E-03 respectively. The RMSE error at 35 °C for different regression algorithms SVR, DT, KNN, and RF are 5.062E-02, 6.862E-03, 6.333E-03, and 1.275E-02 respectively. The MSE error percentages for the testing sample at 35 °C for different regression algorithms are 0.37 %, 0.01 %, 0.01 %, and 0.02 % respectively. The results indicate that KNN and DT exhibit higher precision when applied to lithium-ion batteries for electric vehicles. Collaborating with hardware manufacturers to exchange data has the potential to foster the creation of innovative battery health monitoring systems.

Mobile Customer Satisfaction Scoring Research Based on Quadratic Dimension Reduction and Machine Learning Integration

Customer satisfaction is a measure of the degree of satisfaction of customer experience. Among the three major operators in China, China Mobile plays an important role in the communication field. A study of customer satisfaction with China Mobile will have a significant positive impact on the sustainable development of the entire communication industry. In order to respond to customer needs accurately, a mobile customer satisfaction research method based on quadratic dimensionality reduction and machine learning integration is proposed. Firstly, the core evaluation system of impact satisfaction is established, through the integration of systematic clustering and exploratory factor analysis for quadratic dimensionality reduction. Then, unreasonable data in the core influencing factors are eliminated. Finally, the gradient-boosted decision tree (GBDT) machine learning algorithm is applied to predict satisfaction, with a prediction accuracy of up to 99%, and the highly accurate satisfaction prediction can quickly respond to customer needs and feedback to improve customer experience and satisfaction.

Habitat in flames: How climate change will affect fire risk across koala forests

Aim:Generate fire susceptibility maps for the present and 2070, to identify the threat wildfires pose to koalas now and under future climate change. Location:Australia. Time period:Present and 2070. Major taxa studied:60 main tree species browsed by koalas. Method:The Decision Tree machine learning algorithm was applied to generate a fire susceptibility index (a measure of the potential for a given area or region to experience wildfires) using a dataset of conditioning factors, namely: altitude, aspect, rainfall, distance from rivers, distance from roads, forest type, geology, koala presence and future dietary sources, land use-land cover (LULC), normalized difference vegetation index (NDVI), slope, soil, temperature, and wind speed. Results:We found a general increase in susceptibility of Australian vegetation to bushfires overall. The simulation for current conditions indicated that 39.56% of total koala habitat has a fire susceptibility rating of “very high” or “high”, increasing to 44.61% by 2070. Main conclusions:Wildfires will increasingly impact koala populations in the future. If this iconic and vulnerable marsupial is to be protected, conservation strategies need to be adapted to deal with this threat. It is crucial to strike a balance between ensuring that koala habitats and populations are not completely destroyed by fire while also allowing for forest rejuvenation and regeneration through periodic burns.

A Pseudorange-Based GPS Spoofing Detection Using Hyperbola Equations

Spoofing attacks are one of the most critical threats against secure global navigation satellite system (GNSS) positioning. Since correct positioning is a must in many vehicle-to-everything (V2X) communication systems, the detection of these attacks is vital. In the literature on the detection strategies of spoofing attacks, the majority of the solutions are based on the assumption that all available signals are spoofed. In this paper, we focus on detecting spoofing attacks in which authentic and spoofing positioning signals coexist in a V2X system. The observable pseudorange values are utilized with the help of derived hyperbola equations during the design of the spoofing detection algorithms. We propose an algorithm, which is named as the sub-optimal search-based spoofing detection algorithm (Algorithm 1), and it considers all possible numbers of spoofing attacking signals, but not all spoofing scenarios with the same amount of spoofing signals. To address the complexity problems based on the increased number of search scenarios of this approach, we propose another algorithm, which is called subset selection-based spoofing detection algorithm (Algorithm 2), with a smart selection of the search subsets. Both of these algorithms are first compared with fixed detection thresholds, which are determined with the Pareto front approach. Then, the performance of the algorithms is investigated when vehicle mobility and spoofing imperfection are considered. Finally, a supervised learning-based decision tree machine learning (ML) algorithm is run without specifying any detection threshold. The results indicate that Algorithm 1 provides higher detection rates than the subset selection-based algorithm; however, the false alarm ratios of Algorithm 2 are much lower than its original performance.

Decision Tree Models and Machine Learning Algorithms in the Fault Recognition on Power Lines with Branches

The complication of the structure, topology and composition of the future electrical networks is characterized by difficult-to-recognize circuit-mode situations and requires modern methods for analyzing information parameters. The growing trend of digitizing signals in substations and the use of the IEC 61850 standard results in a huge amount of new data available at the nodes of the electrical network. The development and analysis of new methods for detecting and recognizing the modes of electrical networks (normal and emergency) are topical research issues. The article explores a new approach to recognizing a faulted section of an electrical network with branches by concurrently analyzing several information features and applying machine learning methods: decision tree, random forest, and gradient boosting. The application of this approach for decision-making by relay protection has not been previously implemented. Simulation modeling and the Monte Carlo method are at the heart of obtaining training samples. The results of testing the studied methods under review showed the required flexibility, the ability to use a large number of information parameters, as well as the best results of fault recognition in comparison with the distance protection relay.

Geometric graph learning with extended atom-types features for protein-ligand binding affinity prediction

Understanding and accurately predicting protein-ligand binding affinity are essential in the drug design and discovery process. At present, machine learning-based methodologies are gaining popularity as a means of predicting binding affinity due to their efficiency and accuracy, as well as the increasing availability of structural and binding affinity data for protein-ligand complexes. In biomolecular studies, graph theory has been widely applied since graphs can be used to model molecules or molecular complexes in a natural manner. In the present work, we upgrade the graph-based learners for the study of protein-ligand interactions by integrating extensive atom types such as SYBYL and extended connectivity interactive features (ECIF) into multiscale weighted colored graphs (MWCG). By pairing with the gradient boosting decision tree (GBDT) machine learning algorithm, our approach results in two different methods, namely sybylGGL-Score and ecifGGL-Score. Both of our models are extensively validated in their scoring power using three commonly used benchmark datasets in the drug design area, namely CASF-2007, CASF-2013, and CASF-2016. The performance of our best model sybylGGL-Score is compared with other state-of-the-art models in the binding affinity prediction for each benchmark. While both of our models achieve state-of-the-art results, the SYBYL atom-type model sybylGGL-Score outperforms other methods by a wide margin in all benchmarks. Finally, the best-performing SYBYL atom-type model is evaluated on two test sets that are independent of CASF benchmarks.

FPGA-based fault analysis for 7-level switched ladder multi-level inverter using decision tree algorithm

The proposed method involves the fault analysis of the inverter switches present in the multi-level inverter (MLI) circuitry. The decision tree machine learning algorithm is incorporated for the fault analysis of the inverter switches. The multi-level inverter utilized in this work is a 7-level switched ladder multi-level inverter. There is 4 number of switches in the design of a 7-level inverter driven by the non-carrier digital pulse width modulation signals. The non-carried-based digital pulse-width modulator (DPWM) generation is generated using the event angle for the 7-level of the switched ladder inverter. The proposed method investigates the stuck-at-fault occurrences of the 4 switches in the inverter by manipulating the decision tree parameters such as entropy, information gain, and decision tree. Based on the decision tree, the very high-speed integrated circuit hardware description language (VHDL) code is developed by making use of the behavioral modeling and validated for the power, area in the Xilinx Vivado tool. The real-time feasibility is verified for the proposed method by synthesizing the developed VHDL code in the field programmable gate array (FPGA) device.

Using a decision tree algorithm to distinguish between repeated supra-therapeutic and acute acetaminophen exposures

BackgroundThis study aimed to compare clinical and laboratory characteristics of supra-therapeutic (RSTI) and acute acetaminophen exposures using a predictive decision tree (DT) algorithm.MethodsWe conducted a retrospective cohort study using the National Poison Data System (NPDS). All patients with RSTI acetaminophen exposure (n = 4,522) between January 2012 and December 2017 were included. Additionally, 4,522 randomly selected acute acetaminophen ingestion cases were included. After that, the DT machine learning algorithm was applied to differentiate acute acetaminophen exposure from supratherapeutic exposures.ResultsThe DT model had accuracy, precision, recall, and F1-scores of 0.75, respectively. Age was the most relevant variable in predicting the type of acetaminophen exposure, whether RSTI or acute. Serum aminotransferase concentrations, abdominal pain, drowsiness/lethargy, and nausea/vomiting were the other most important factors distinguishing between RST and acute acetaminophen exposure.ConclusionDT models can potentially aid in distinguishing between acute and RSTI of acetaminophen. Further validation is needed to assess the clinical utility of this model.

Early Childhood Predictors for Dental Caries: A Machine Learning Approach

We aimed to develop and validate caries prognosis models in primary and permanent teeth after 2 and 10 y of follow-up through a machine learning (ML) approach, using predictors collected in early childhood. Data from a 10-y prospective cohort study conducted in southern Brazil were analyzed. Children aged 1 to 5 y were first examined in 2010 and reassessed in 2012 and 2020 regarding caries development. Dental caries was assessed using the Caries Detection and Assessment System (ICDAS) criteria. Demographic, socioeconomic, psychosocial, behavioral, and clinical factors were collected. ML algorithms decision tree, random forest, and extreme gradient boosting (XGBoost) were employed, along with logistic regression. The discrimination and calibration of models were verified in independent sets. From 639 children included at the baseline, we reassessed 467 (73.3%) and 428 (66.9%) children in 2012 and 2020, respectively. For all models, the area under receiver operating characteristic curve (AUC) at training and testing was above 0.70 for predicting caries in primary teeth after 2-y follow-up, with caries severity at the baseline being the strongest predictor. After 10 y, the SHAP algorithm based on XGBoost achieved an AUC higher than 0.70 in the testing set and indicated caries experience, nonuse of fluoridated toothpaste, parent education, higher frequency of sugar consumption, low frequency of visits to the relatives, and poor parents’ perception of their children’s oral health as top predictors for caries in permanent teeth. In conclusion, the implementation of ML shows potential for determining caries development in both primary and permanent teeth using easy-to-collect predictors in early childhood.

Decision trees, entropy, and the contrastive feature hierarchy

Dresher (2009) argues that language-particular hierarchies of features are the best way to identify contrastive features in a phonological inventory. While not universal, this ordering of features is also not fully unconstrained. But what limits the space of possible feature orders remains an open question. This paper demonstrates how the concept of entropy establishes a partial ordering of features that both allows for but also constrains language-particular variation. Specifically, a decision tree machine learning algorithm is employed to dynamically impose structure on the hypothesis space of possible feature orders.

Machine learning for design of degenerate Cas13a crRNAs using lassa virus as a model of highly variable RNA target

The design of minimum CRISPR RNA (crRNA) sets for detection of diverse RNA targets using sequence degeneracy has not been systematically addressed. We tested candidate degenerate Cas13a crRNA sets designed for detection of diverse RNA targets (Lassa virus). A decision tree machine learning (ML) algorithm (RuleFit) was applied to define the top attributes that determine the specificity of degenerate crRNAs to elicit collateral nuclease activity. Although the total number of mismatches (0–4) is important, the specificity depends as well on the spacing of mismatches, and their proximity to the 5’ end of the spacer. We developed a predictive algorithm for design of candidate degenerate crRNA sets, allowing improved discrimination between “included” and “excluded” groups of related target sequences. A single degenerate crRNA set adhering to these rules detected representatives of all Lassa lineages. Our general ML approach may be applied to the design of degenerate crRNA sets for any CRISPR/Cas system.

Technology for Kisan Samanvayam: Nutrition Intelligibility of Groundnut Plant using IoT-ML Framework

Neolithic Demographic transition resulting the reduction of habitable land for cultivation. Hence the smart agriculture is the only way to cater higher food demand. The farming community of developing countries like India needs Kisan Samanvayam with futuristic technologies for financially viable cultivation. Technology place vital role in economically nourishment of soil fertility and crop management. In this regard we proposed IoT-ML framework for remotely assessing the soil nutrients (N, P,K), PH and early stage detection of crop deceases. Android APP which is a part and parcel of the frame work enable the farmer to have real time visual statistics of the soil nutrients, notifications and suggestions regarding to the crop management. JXCT Soil NPK sensors, PH sensors, Dual Core ESP32 Controllers, Firebase Cloud and Random Forest Decision Tree machine Learning Algorithm, Micromlgen serve this purpose. Unlike Solitary sensor for entire field, we have divided a hector into four subregions for effective monitoring local region needs. The presence of IoT with TinyML increased the robustness of the framework and results are encouraging with sandy loam soil.

Sentiment Analysis of Covid-19 Handling in Indonesia Based on Lexicon Weighting

Covid-19, a contagious disease, has been classified as a global pandemic. Indonesia, as one of the ASEAN countries, has taken various measures to combat the spread of this disease. One of the government's initiatives to tackle the pandemic is the PeduliLindungi application, through which the public provides feedback on government policies. However, analyzing and comprehending public opinions in a non-subjective manner poses a challenge in objectively evaluating government services. This study aims to address this issue by conducting a sentiment analysis of Covid-19 handling in Indonesia, using a lexicon-based weighting system that includes SentiStrengthID and InSet. The decision tree (DT) machine learning algorithm is utilized to evaluate the polarity results provided by the lexicon. The results indicate that the sentiment polarity towards Covid-19 handling in Indonesia is negative based on both SentiStrengthID and InSet weights. Evaluating machine learning performance with the SentiStrengthID lexicon, the DT-entropy and DT-gini models achieved an accuracy of 82% and 83%, respectively. Similarly, evaluating machine learning performance with the InSet lexicon, the DT-entropy and DT-gini models achieved an accuracy of 81% and 82%, respectively.