Patterns In Complex Datasets Research Articles

Latent personality profiles of analog astronauts: An unsupervised clustering method analysis

The selection of astronauts for space missions is an important process that involves the assessment of multiple factors, including personality traits. Previous studies have corroborated that these traits exert a significant influence on the performance and adaptation of the crew during the execution of space missions. In this study, International Personality Item Pool - NEO – 120 (IPIP-120) was used to assess the personality of participants in a sample of 43 analog astronauts who took part in 9 different analog missions between 2022 and 2023. The objective was to identify specific personality profiles among the analog astronauts using the unsupervised k-means clustering method. The centroids of each cluster were analyzed to describe distinctive personality profiles. The results revealed the existence of two distinct personality clusters among the analog astronauts. The first cluster showed a greater propensity toward adventure, intellect, goal striving, extroversion, and cooperation. In contrast, the second cluster exhibited lower scores on these traits. The identification of these two differentiated personality profiles reinforces the importance of personality assessment as a valuable tool in the astronaut selection process. Furthermore, the application of the k-means method demonstrates its utility as an effective technique for identifying underlying patterns in complex datasets. These findings may inform future astronaut selection strategies, contributing to the formation of more effective and cohesive teams for space missions.

Sociodemographic Variables in Offender and Non-Offender Patients Diagnosed with Schizophrenia Spectrum Disorders-An Explorative Analysis Using Machine Learning.

With the growing availability of medical data and the enhanced performance of computers, new opportunities for data analysis in research are emerging. One of these modern approaches is machine learning (ML), an advanced form of statistics broadly defined as the application of complex algorithms. ML provides innovative methods for detecting patterns in complex datasets. This enables the identification of correlations or the prediction of specific events. These capabilities are especially valuable for multifactorial phenomena, such as those found in mental health and forensic psychiatry. ML also allows for the quantification of the quality of the emerging statistical model. The present study aims to examine various sociodemographic variables in order to detect differences in a sample of 370 offender patients and 370 non-offender patients, all with schizophrenia spectrum disorders, through discriminative model building using ML. In total, 48 variables were tested. Out of seven algorithms, gradient boosting emerged as the most suitable for the dataset. The discriminative model finally included three variables (regarding country of birth, residence status, and educational status) and yielded an area under the curve (AUC) of 0.65, meaning that the statistical discrimination of offender and non-offender patients based purely on the sociodemographic variables is rather poor.

Leveraging Machine Learning Approaches to Predict Organic Carbon Abundance in Mars‐Analog Hypersaline Lake Sediments

AbstractModern advancements in laboratory and instrumental techniques in astrobiology have improved our life detection capabilities on both Earth and beyond. These advancements have also increased the complexity of data often resulting in data sets that are characterized by complex and non‐linear relationships. Machine learning methods are underutilized in astrobiology; however, these methods are extremely effective at revealing structure and patterns in complex data sets when paired with the right algorithms. Here, we employ a series of classification and regression algorithms to predict the abundance of organic carbon (OC) from X‐ray fluorescence (XRF) heavy element (>Mg) data in dynamic Mars‐analog hypersaline lake sediments. More specifically, we constructed models using the random forest, k‐nearest neighbors (KNN), support vector machine, and logistic regression algorithms. Overall, our trained models showed good performance with predicting the abundance of OC, with accuracies from 80% to 94%. Machine learning approaches such as classification and regression algorithms offer insight into complex data while providing agnostic insights, ultimately creating a more efficient search for OC. We applied our trained model on XRF data from Martian soil using rover‐based (PIXL) and orbital (Odyssey) data sets to produce probability predictions of OC abundance. Our predictions show a high probability that OC abundance is low which is comparable to OC data from recently landed missions. These results highlight the potential for predictive machine learning models to be trained on data from analog environments on Earth and then applied to extraterrestrial targets, ultimately, improving life detection efforts.

Meta Insights: Analyzing Hero Performance in Mobile Legends with K-Nearest Neighbors

This research presents a thorough statistical analysis of hero performance in the latest Mobile Legends meta, employing the K-Nearest Neighbor (KNN) algorithm. Utilizing diverse data sources, the study explores factors influencing hero success, leveraging KNN's ability to identify intricate patterns in complex datasets. Through meticulous data collection, preprocessing, and application of the KNN algorithm, the research classifies and predicts hero performance based on similarities with neighboring heroes. Critical determinants such as win rates, popularity, and hero ban rates emerge, providing profound insights into gameplay strategies. The study emphasizes the importance of understanding meta dynamics, hero attributes, and player expertise for informed decision-making in hero selection within the dynamic landscape of Mobile Legends.

Unsupervised Learning and Pattern Recognition in Alloy Design

Machine learning has the potential to revolutionise alloy design by uncovering useful patterns in complex datasets and supplementing human expertise and experience. This review examines the role of unsupervised learning...

A Comparative Analysis Of Logistic Regression and K-Nearest Neighbors Algorithms In Diagnosis Of Diabetes

Machine Learning techniques have gained prominence in medical diagnosis due to their ability to uncover patterns in complex data-sets, thereby giving accurate disease classification. In this study, we mainly focus on the application of two widely used Machine Learning algorithms, Logistic Regression and K-Nearestneighbors( KNN), for the purpose of distinguishing patients with diabetes from those without. Our research aims to shed light on the comparative accuracy and performance of these algorithms in a medical context. The methodology section outlines experimental setup, detailing data processing, algorithm training and testing procedures. A comprehensive data-set comprising medical attributes is utilized for evaluation and accuracy metrics are employed to quantify the performance of the algorithms. Results has shown efficacy of both the algorithms and our findings showcase the strengths and limitations of each approach, contributing on the applicability in medical decision making. By offering a nuanced comparison, we illuminate a path for more robust and accurate disease identification techniques, further enhancing patient care and medical outcomes. GANIT J. Bangladesh Math. Soc. 43.1 (2023) 01- 07

A Comparison of Three Machine Learning Methods for Multivariate Genomic Prediction Using the Sparse Kernels Method (SKM) Library.

Genomic selection (GS) changed the way plant breeders select genotypes. GS takes advantage of phenotypic and genotypic information to training a statistical machine learning model, which is used to predict phenotypic (or breeding) values of new lines for which only genotypic information is available. Therefore, many statistical machine learning methods have been proposed for this task. Multi-trait (MT) genomic prediction models take advantage of correlated traits to improve prediction accuracy. Therefore, some multivariate statistical machine learning methods are popular for GS. In this paper, we compare the prediction performance of three MT methods: the MT genomic best linear unbiased predictor (GBLUP), the MT partial least squares (PLS) and the multi-trait random forest (RF) methods. Benchmarking was performed with six real datasets. We found that the three investigated methods produce similar results, but under predictors with genotype (G) and environment (E), that is, E + G, the MT GBLUP achieved superior performance, whereas under predictors E + G + genotype environment (GE) and G + GE, random forest achieved the best results. We also found that the best predictions were achieved under the predictors E + G and E + G + GE. Here, we also provide the R code for the implementation of these three statistical machine learning methods in the sparse kernel method (SKM) library, which offers not only options for single-trait prediction with various statistical machine learning methods but also some options for MT predictions that can help to capture improved complex patterns in datasets that are common in genomic selection.

Defining resistance mechanisms to CDK4/6 inhibition in hormone receptor-positive HER2-negative metastatic breast cancer (MBC) through a machine learning approach applied to circulating tumor DNA (ctDNA).

3055 Background: Although cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) are a primary treatment for hormone receptor-positive/HER2 negative MBC, data regarding resistance mechanisms are still an unmet need. The aim of the study was to highlight new resistance pathways using machine learning (ML) to detect multiparametric patterns in complex datasets. Methods: The study retrospectively analyzed a cohort of 610 hormone receptor positive HER2 negative MBC patients (pts) at Northwestern University, Massachusetts General Hospital and Washington University in St. Louis between 2015-2020 with baseline ctDNA testing by Guardant360. Pathways were defined based on previous work (Sanchez-Vega F et al, Cell. 2018) (i.e., RTK, RAS, RAF, MEK, NRF2, ER, WNT, MYC, P53, cell cycle, notch, PI3K). Only pathogenic variants according to OncoKB were included in the models. Associations among single nucleotide (SNVs) and copy number (CNVs) variations, pathway classification and previous exposure to CDK4/6i were explored through logistic regression and Gradient boosted machines (GBMs) ML algorithm. Results: at baseline, 322 pts (52.8%) were previously treated with CDK4/6i. The most detected pathway alterations were SNVs in PI3K (37.1%), P53 (31.8%), ER (29.2%) and RTK (22.3%). After stepwise logistic regression, RB1, NF1 and ESR1 SNVs were associated with previous exposure to CDK4/6i (respectively OR: 3.55 P = 0.017; OR: 3.06 P = 0.026 and OR: 1.82 P < 0.001), while SNVs in the ER pathway were associated with CDK4/6i (1.56 P < 0.001). Two GBMs models were designed based on gene variants (training AUC: 0.695, cross validation AUC: 0.631) and oncogenic pathways (training AUC: 0.713, cross validation AUC: 0.619). The highest relative importance (RI) was observed for ESR1 SNVs (RI: 35.35), TP53 SNVs (RI: 11.33), NF1 SNVs (RI: 3.45), SMAD4 SNVs (RI: 3.39) and RB1 SNVs (RI: 3.33). Alterations at a pathway level with the highest RI were ER SNVs (RI: 33.50), P53 SNVs (RI: 14.98), PI3K SNVs (RI: 14.40), RTK SNVs (RI: 10.55), RTK CNVs (RI: 10.26), cell cycle CNVs (RI: 6.99), cell cycle SNVs (RI: 6.77) and RAS SNVs (RI: 6.54). Of the previously highlighted pathway alterations, a significant impact on PFS after ctDNA collection was observed among de novo pts treated with CDK4/6i (165 pts) for ER SNVs (P < 0.0001), RTK SNVs (P = 0.0011), RTK CNVs (P = 0.0006), Cell cycle CNVs (P = 0.0010) and Cell cycle SNVs (P = 0.0143). No impact was observed on PFS for pts who had not received a CDK4/6i-based regimen. Conclusions: The combination of ctDNA-based datasets and machine learning algorithms defined novel resistance pathways for patients treated with CDK4/6i. Although preliminary, these results suggest that alterations of the ER, RTK and Cell cycle pathways might be crucial to optimize treatment strategy and drug development.

Using Machine Learning to Generate Storm-Scale Probabilistic Guidance of Severe Weather Hazards in the Warn-on-Forecast System

AbstractA primary goal of the National Oceanic and Atmospheric Administration Warn-on-Forecast (WoF) project is to provide rapidly updating probabilistic guidance to human forecasters for short-term (e.g., 0–3 h) severe weather forecasts. Postprocessing is required to maximize the usefulness of probabilistic guidance from an ensemble of convection-allowing model forecasts. Machine learning (ML) models have become popular methods for postprocessing severe weather guidance since they can leverage numerous variables to discover useful patterns in complex datasets. In this study, we develop and evaluate a series of ML models to produce calibrated, probabilistic severe weather guidance from WoF System (WoFS) output. Our dataset includes WoFS ensemble forecasts available every 5 min out to 150 min of lead time from the 2017–19 NOAA Hazardous Weather Testbed Spring Forecasting Experiments (81 dates). Using a novel ensemble storm-track identification method, we extracted three sets of predictors from the WoFS forecasts: intrastorm state variables, near-storm environment variables, and morphological attributes of the ensemble storm tracks. We then trained random forests, gradient-boosted trees, and logistic regression algorithms to predict which WoFS 30-min ensemble storm tracks will overlap a tornado, severe hail, and/or severe wind report. To provide rigorous baselines against which to evaluate the skill of the ML models, we extracted the ensemble probabilities of hazard-relevant WoFS variables exceeding tuned thresholds from each ensemble storm track. The three ML algorithms discriminated well for all three hazards and produced more reliable probabilities than the baseline predictions. Overall, the results suggest that ML-based postprocessing of dynamical ensemble output can improve short-term, storm-scale severe weather probabilistic guidance.

Disentangling etiologies of CNS infections in Singapore using multiple correspondence analysis and random forest

Central nervous system (CNS) infections cause substantial morbidity and mortality worldwide, with mounting concern about new and emerging neurologic infections. Stratifying etiologies based on initial clinical and laboratory data would facilitate etiology-based treatment rather than relying on empirical treatment. Here, we report the epidemiology and clinical outcomes of patients with CNS infections from a prospective surveillance study that took place between 2013 and 2016 in Singapore. Using multiple correspondence analysis and random forest, we analyzed the link between clinical presentation, laboratory results, outcome and etiology. Of 199 patients, etiology was identified as infectious in 110 (55.3%, 95%-CI 48.3–62.0), immune-mediated in 10 (5.0%, 95%-CI 2.8–9.0), and unknown in 79 patients (39.7%, 95%-CI 33.2–46.6). The initial presenting clinical features were associated with the prognosis at 2 weeks, while laboratory-related parameters were related to the etiology of CNS disease. The parameters measured were helpful to stratify etiologies in broad categories, but were not able to discriminate completely between all the etiologies. Our results suggest that while prognosis of CNS is clearly related to the initial clinical presentation, pinpointing etiology remains challenging. Bio-computational methods which identify patterns in complex datasets may help to supplement CNS infection diagnostic and prognostic decisions.

Machine learning in plant-pathogen interactions: empowering biological predictions from field scale to genome scale.

Machine learning (ML) encompasses statistical methods that learn to identify patterns in complex datasets. Here, I review application areas in plant-pathogen interactions that have recently benefited from ML, such as disease monitoring, the discovery of gene regulatory networks, genomic selection for disease resistance and prediction of pathogen effectors. However, achieving robust performance from ML is not trivial and requires knowledge of both the methodology and the biology. I discuss common pitfalls and challenges in using ML approaches. Finally, I highlight future opportunities for ML as a tool for dissecting plant-pathogen interactions using high-throughput data, for example, through integration of diverse data sources and the analysis with higher resolution, such as from individual cells or on elaborate spatial and temporal scales.

The Use of Artificial Neural Networks in Projectile Point Typology

Regional empirical typologies have received criticisms for characterizing artifact types as immutable, mutually exclusive groups, rather than analytical groups developed for an analytical purposes. The research presented here uses an empirical regional typology to supply data for the creation of analytical units. Developed in the field of computer science as a means to discover complex patterns in data sets, archaeologists can use artificial neural networks to identify analytical patterns in typological data. In this paper we explain what neural networks are and how they work. We show how a specific architecture and training process can be applied to artificial neural networks to create analytical units from empirical units. Finally, we provide four examples to show that the networks will not find a pattern where none exists, and to demonstrate how artificial neural networks can be applied to solve more complex problems.

Systems Biology and Addiction

The onset of addiction is marked with drug induced positive experiences that keep being repeated. During that time, adaptation occurs and addiction is stabilized. Interruption of those processes induces polysymptomatic withdrawal syndromes. Abstinence is accompanied by risks of relapse. These features of addiction suggest adaptive brain dynamics with common pathways in complex neuronal networks. Addiction research has used animal models, where some of those phenomena could be reproduced, to find correlates of addictive behavior. The major thrust of those approaches has been on the involvement of genes and proteins. Recently, an enormous amount of data has been obtained by high throughput technologies in these fields. Therefore, (Computational) "Systems Biology" had to be implemented as a new approach in molecular biology and biochemistry. Conceptually, Systems Biology can be understood as a field of theoretical biology that tries to identify patterns in complex data sets and that reconstructs the cell and cellular networks as complex dynamic, self-organizing systems. This approach is embedded in systems science as an interdisciplinary effort to understand complex dynamical systems and belongs to the field of theoretical neuroscience (Computational Neuroscience). Systems biology, in a similar way as computational neuroscience is based on applied mathematics, computer-based computation and experimental simulation. In terms of addiction research, building up "computational molecular systems biology of the (addicted) neuron" could provide a better molecular biological understanding of addiction on the cellular and network level. Some key issues are addressed in this article.

Exploring Cognitive Strategies for Integrating Multiple-View Visualizations

Visualizing information is useful for finding patterns in complex data sets, but little research has been done on how people understand multiple-view visualizations (multiple visualizations presented simultaneously). A controlled experiment was performed using different combinations of visualizations and different task types as independent variables, and qualitative and quantitative data were collected. To collect the data psychological tests, logs of the participants' interaction, eye-tracking equipment, and video recordings were used. This paper reports a portion of the results from this experiment. Main findings include that, contrary to what was suggested in previous literature, the time cost for switching between different types of visualizations (context switching) may not be significant, and that displaying the data using the same type of visualization may cause interference. Orthogonal combinations appear to aid users in finding and recognizing patterns, and focusing attention and analogical reasoning on spatial relationships may be important cognitive abilities for the given tasks.

Artificial neural networks: a potential role in osteoporosis.

Artificial neural networks are computer software systems that recognize patterns in complex data sets. A recent development in neural computing, multiversion systems (MVS), has led to enhanced analytical power, and this was harnessed to demonstrate the value of risk factors in predicting the result of osteoporosis investigations by quantitative ultrasound. 274 women were screened in an open-access osteoporosis service. A conventional risk factor questionnaire was completed for each patient by the osteoporosis specialist nurse. An MVS was trained on 180 randomly selected data sets and tested on the remaining 94. The results were compared with those from logistic regression analysis in predictive power, both from the selected 20-item questionnaire and for a limited 5-item questionnaire comprising age, height, height loss, weight and years since the menopause. The MVS approach predicted the T-score categorization of the patients from the 20-item questionnaire with 83.0% accuracy, whereas logistic regression yielded an accuracy of only 72.8% (P = 0.04). From the 5-item database the MVS yielded a best prediction accuracy of 73.1%, whereas the logistic regression prediction accuracy was 60% (P = 0.04). These results suggest that 20 risk factors can be used by an MVS to predict the outcome of osteoporosis investigations with a power that outperforms conventional statistical methods. Use of this system may improve the selection of patients for osteoporosis investigations, since even with only 5 risk factors the system performs nearly as well as that based on the full 20 factors.

Detrended Correspondence Analysis of Phytoplankton Abundance and Distribution in Sandusky Bay and Lake Erie

The purpose of this study was to determine if phytoplankton communities in Sandusky Bay were distinct from those of Lake Erie. Samples were taken from 11 sites along a 50 km transect extending from the lower reaches of the Sandusky River, through Sandusky Bay and into the western basin of Lake Erie to identify factors correlated with identifiable patterns in distribution and abundance of summer phytoplankton. Detrended correspondence analysis (DCA), an ordination technique used to describe patterns in complex data sets, arranged the sample sites along an ordination axis that explained 76% of the variation in the phytoplankton abundance data matrix, and produced the following sequence of ordination axis scores: Sandusky Bay→ Lake Erie→ Sandusky River. DCA axis I scores strongly correlated with total phosphorus, soluble reactive phosphorus, algal phosphatase activity, dissolved oxygen, conductivity, turbidity, and alkalinity, but not chloride concentration, suggesting that phytoplankton abundance and distribution were related to phosphorus availability and not simply due to the passive movement of water along the transect. Bacterial abundance correlated with DCA axis I, suggesting that phytoplankton-bacterial interactions may be important in understanding distributional patterns of Sandusky Bay phytoplankton.

The use of lichens in atmospheric deposition studies with an emphasis on the Arctic

Lichens are well known to accumulate and retain a variety of contaminants and therefore are very useful in documenting relative temporal and spatial deposition patterns of contaminants. As one of the dominant and most widespread groups of arctic plants, they have been used frequently in deposition studies, particularly as related to redionuclides and metals. Many lichens are highly dependent on atmospheric sources of nutrients. It is assumed that deposition and accumulation processes related to contaminants occur in ways analogous to nutrient deposition and accumulation. As receptors of atmospheric deposition, lichens have advantages over vascular plants due to the lichens' relative lack of morphological variability with time, their lack of a cuticle and associated stomates and their lack of roots. Mechanisms by which contaminants are accumulated and tolerated by lichens are briefly reviewed. Data regarding elevated elemental concentrations associated with certain industrial activities in relation to putative background levels are tabulated. The use of enrichment factors and multivariate analyses provide two procedures by which patterns in complex data sets can be discerned where pollution is minimal. Such analyses are presented as a case study from the Northwest Territories of Canada. Finally, arctic lichens may also be useful in documenting deposition patterns of persistent organic contaminants, particularly in relation to the lichen-caribou-human food chain, as was previously documented for radionuclides.