scholarly journals Semantic similarity and machine learning with ontologies

2020 ◽  
Author(s):  
Maxat Kulmanov ◽  
Fatima Zohra Smaili ◽  
Xin Gao ◽  
Robert Hoehndorf
Keyword(s):  
Machine Learning ◽  
Semantic Similarity ◽  
Domain Knowledge ◽  
Life Sciences ◽  
Similarity Measures ◽  
Background Knowledge ◽  
Biological Database ◽  
Learning Models ◽  
Machine Learning Methods ◽  
Machine Learning Models

Abstract Ontologies have long been employed in the life sciences to formally represent and reason over domain knowledge and they are employed in almost every major biological database. Recently, ontologies are increasingly being used to provide background knowledge in similarity-based analysis and machine learning models. The methods employed to combine ontologies and machine learning are still novel and actively being developed. We provide an overview over the methods that use ontologies to compute similarity and incorporate them in machine learning methods; in particular, we outline how semantic similarity measures and ontology embeddings can exploit the background knowledge in ontologies and how ontologies can provide constraints that improve machine learning models. The methods and experiments we describe are available as a set of executable notebooks, and we also provide a set of slides and additional resources at https://github.com/bio-ontology-research-group/machine-learning-with-ontologies.

scholarly journals Machine learning with biomedical ontologies

2020 ◽  
Author(s):  
Maxat Kulmanov ◽  
Fatima Zohra Smaili ◽  
Xin Gao ◽  
Robert Hoehndorf
Keyword(s):  
Machine Learning ◽  
Research Group ◽  
Domain Knowledge ◽  
Background Knowledge ◽  
Search Space ◽  
Biological Database ◽  
List Type ◽  
Biomedical Ontologies ◽  
Learning Models ◽  
Machine Learning Models

Ontologies have long been employed in the life sciences to formally represent and reason over domain knowledge, and they are employed in almost every major biological database. Recently, ontologies are increasingly being used to provide background knowledge in similarity-based analysis and machine learning models. The methods employed to combine ontologies and machine learning are still novel and actively being developed. We provide an overview over the methods that use ontologies to compute similarity and incorporate them in machine learning methods; in particular, we outline how semantic similarity measures and ontology embeddings can exploit the background knowledge in biomedical ontologies, and how ontologies can provide constraints that improve machine learning models. The methods and experiments we describe are available as a set of executable notebooks, and we also provide a set of slides and additional resources at https://github.com/bio-ontology-research-group/machine-learning-with-ontologies.Key pointsOntologies provide background knowledge that can be exploited in machine learning models.Ontology embeddings are structure-preserving maps from ontologies into vector spaces and provide an important method for utilizing ontologies in machine learning. Embeddings can preserve different structures in ontologies, including their graph structures, syntactic regularities, or their model-theoretic semantics.Axioms in ontologies, in particular those involving negation, can be used as constraints in optimization and machine learning to reduce the search space.

Utilizing Blockchain Technology in Social Media Bot Identification

2020 ◽  
Author(s):  
Shreya Reddy ◽  
Lisa Ewen ◽  
Pankti Patel ◽  
Prerak Patel ◽  
Ankit Kundal ◽  
...  
Keyword(s):  
Machine Learning ◽  
Social Media ◽  
Gold Standard ◽  
The Internet ◽  
Learning Models ◽  
Current Time ◽  
Machine Learning Methods ◽  
Blockchain Technology ◽  
Modern Age ◽  
Machine Learning Models

<p>As bots become more prevalent and smarter in the modern age of the internet, it becomes ever more important that they be identified and removed. Recent research has dictated that machine learning methods are accurate and the gold standard of bot identification on social media. Unfortunately, machine learning models do not come without their negative aspects such as lengthy training times, difficult feature selection, and overwhelming pre-processing tasks. To overcome these difficulties, we are proposing a blockchain framework for bot identification. At the current time, it is unknown how this method will perform, but it serves to prove the existence of an overwhelming gap of research under this area.<i></i></p>

scholarly journals Application of Machine Learning Methods in the Task of Identifying User Accounts in Two Social Networks

2019 ◽  
pp. 29-43
Author(s):  
Anastasiya A. Korepanova ◽  
◽  
Valerii D. Oliseenko ◽  
Maxim V. Abramov ◽  
Alexander L. Tulupyev ◽  
...  
Keyword(s):  
Machine Learning ◽  
Social Networks ◽  
Information System ◽  
New Combination ◽  
Practical Significance ◽  
User Profiles ◽  
Learning Models ◽  
Machine Learning Methods ◽  
The Social ◽  
Machine Learning Models

The article describes the approach to solving the problem of comparing user profiles of different social networks and identifying those that belong to one person. An appropriate method is proposed based on a comparison of the social environment and the values of account profile attributes in two different social networks. The results of applying various machine learning models to solving this problem are compared. The novelty of the approach lies in the proposed new combination of various methods and application to new social networks. The practical significance of the study is to automate the process of determining the ownership of profiles in various social networks to one user. These results can be applied in the task of constructing a meta-profile of a user of an information system for the subsequent construction of a profile of his vulnerabilities, as well as in other studies devoted to social networks.

scholarly journals Automated Retraining of Machine Learning Models

2019 ◽  
Vol 8 (12) ◽  
pp. 445-452
Keyword(s):  
Machine Learning ◽  
Input Data ◽  
Research Work ◽  
Learning Models ◽  
Machine Learning Methods ◽  
Machine Learning Model ◽  
Crucial Component ◽  
Conventional Machine ◽  
Over Time ◽  
Machine Learning Models

Data is the most crucial component of a successful ML system. Once a machine learning model is developed, it gets obsolete over time due to presence of new input data being generated every second. In order to keep our predictions accurate we need to find a way to keep our models up to date. Our research work involves finding a mechanism which can retrain the model with new data automatically. This research also involves exploring the possibilities of automating machine learning processes. We started this project by training and testing our model using conventional machine learning methods. The outcome was then compared with the outcome of those experiments conducted using the AutoML methods like TPOT. This helped us in finding an efficient technique to retrain our models. These techniques can be used in areas where people do not deal with the actual working of a ML model but only require the outputs of ML processes

scholarly journals A pitfall for machine learning methods aiming to predict across cell types

2020 ◽  
Vol 21 (1) ◽  
Author(s):  
Jacob Schreiber ◽  
Ritambhara Singh ◽  
Jeffrey Bilmes ◽  
William Stafford Noble
Keyword(s):  
Gene Expression ◽  
Machine Learning ◽  
Cell Types ◽  
Chromatin Domain ◽  
Learning Models ◽  
Machine Learning Methods ◽  
Domain Boundaries ◽  
Average Activity ◽  
Test Sets ◽  
Machine Learning Models

AbstractMachine learning models that predict genomic activity are most useful when they make accurate predictions across cell types. Here, we show that when the training and test sets contain the same genomic loci, the resulting model may falsely appear to perform well by effectively memorizing the average activity associated with each locus across the training cell types. We demonstrate this phenomenon in the context of predicting gene expression and chromatin domain boundaries, and we suggest methods to diagnose and avoid the pitfall. We anticipate that, as more data becomes available, future projects will increasingly risk suffering from this issue.

scholarly journals Sex estimation in cranial remains: A comparison of machine learning and discriminant analysis in Italian populations

2020 ◽  
Author(s):  
A Pozzi ◽  
C Raffone ◽  
MG Belcastro ◽  
TL Camilleri-Carter
Keyword(s):  
Machine Learning ◽  
Discriminant Analysis ◽  
Linear Discriminant Analysis ◽  
Foramen Magnum ◽  
Sex Estimation ◽  
Learning Models ◽  
Linear Discriminant ◽  
Machine Learning Methods ◽  
Cranial Measurements ◽  
Machine Learning Models

ABSTRACTObjectivesUsing cranial measurements in two Italian populations, we compare machine learning methods to the more traditional method of linear discriminant analysis in estimating sex. We use crania in sex estimation because it is useful especially when remains are fragmented or displaced, and the cranium may be the only remains found.Materials and MethodsUsing the machine learning methods of decision tree learning, support-vector machines, k-nearest neighbor algorithm, and ensemble methods we estimate the sex of two populations: Samples from Bologna and samples from the island of Sardinia. We used two datasets, one containing 17 cranial measurements, and one measuring the foramen magnum.Results and DiscussionOur results indicate that machine learning models produce similar results to linear discriminant analysis, but in some cases machine learning produces more consistent accuracy between the sexes. Our study shows that sex can be accurately predicted (> 80%) in Italian populations using the cranial measurements we gathered, except for the foramen magnum, which shows a level of accuracy of ∼70% accurate which is on par with previous geometric morphometrics studies using crania in sex estimation. We also find that our trained machine learning models produce population-specific results; we see that Italian crania are sexually dimorphic, but the features that are important to this dimorphism differ between the populations.

scholarly journals Benchmarking machine learning models for the analysis of genetic data using FRESA.CAD Binary Classification Benchmarking

2019 ◽  
Author(s):  
Javier de Velasco Oriol ◽  
Antonio Martinez-Torteya ◽  
Victor Trevino ◽  
Israel Alanis ◽  
Edgar E. Vallejo ◽  
...  
Keyword(s):  
Machine Learning ◽  
Model Selection ◽  
Binary Classification ◽  
Genetic Data ◽  
R Package ◽  
Learning Models ◽  
Classification Problems ◽  
Machine Learning Methods ◽  
Computational Perspective ◽  
Machine Learning Models

AbstractBackgroundMachine learning models have proven to be useful tools for the analysis of genetic data. However, with the availability of a wide variety of such methods, model selection has become increasingly difficult, both from the human and computational perspective.ResultsWe present the R package FRESA.CAD Binary Classification Benchmarking that performs systematic comparisons between a collection of representative machine learning methods for solving binary classification problems on genetic datasets.ConclusionsFRESA.CAD Binary Benchmarking demonstrates to be a useful tool over a variety of binary classification problems comprising the analysis of genetic data showing both quantitative and qualitative advantages over similar packages.

Improved Interpretability of Machine Learning Model Using Unsupervised Clustering: Predicting Time to First Treatment in Chronic Lymphocytic Leukemia

2019 ◽  
pp. 1-11 ◽  
Author(s):  
David Chen ◽  
Gaurav Goyal ◽  
Ronald S. Go ◽  
Sameer A. Parikh ◽  
Che G. Ngufor
Keyword(s):  
Machine Learning ◽  
Risk Stratification ◽  
Unsupervised Clustering ◽  
Support Vector ◽  
Learning Models ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Continuous Output ◽  
Time To First Treatment ◽  
Machine Learning Models

PURPOSE Time to event is an important aspect of clinical decision making. This is particularly true when diseases have highly heterogeneous presentations and prognoses, as in chronic lymphocytic lymphoma (CLL). Although machine learning methods can readily learn complex nonlinear relationships, many methods are criticized as inadequate because of limited interpretability. We propose using unsupervised clustering of the continuous output of machine learning models to provide discrete risk stratification for predicting time to first treatment in a cohort of patients with CLL. PATIENTS AND METHODS A total of 737 treatment-naïve patients with CLL diagnosed at Mayo Clinic were included in this study. We compared predictive abilities for two survival models (Cox proportional hazards and random survival forest) and four classification methods (logistic regression, support vector machines, random forest, and gradient boosting machine). Probability of treatment was then stratified. RESULTS Machine learning methods did not yield significantly more accurate predictions of time to first treatment. However, automated risk stratification provided by clustering was able to better differentiate patients who were at risk for treatment within 1 year than models developed using standard survival analysis techniques. CONCLUSION Clustering the posterior probabilities of machine learning models provides a way to better interpret machine learning models.

P1925Machine learning versus classic electrocardiographic criteria for left ventricular hypertrophy in a young pre-participation cohort: results from the SAFE protocol study

2019 ◽  
Vol 40 (Supplement_1) ◽  
Author(s):  
G Sng ◽  
D Y Z Lim ◽  
C H Sia ◽  
J S W Lee ◽  
X Y Shen ◽  
...  
Keyword(s):  
Machine Learning ◽  
Left Ventricular Hypertrophy ◽  
Ventricular Hypertrophy ◽  
Left Ventricular ◽  
Superior Performance ◽  
Learning Models ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Ecg Criteria ◽  
Machine Learning Models

Abstract Background/Introduction Classic electrocardiographic (ECG) criteria for left ventricular hypertrophy (LVH) have been well studied in Western populations, particularly in hypertensive patients. However, their utility in Asian populations is not well studied, and their applicability to young pre-participation cohorts is unclear. We sought to evaluate the performance of classical criteria against that of machine learning models. Aims We sought to evaluate the performance of classical criteria against the performance of novel machine learning models in the identification of LVH. Methodology Between November 2009 and December 2014, pre-participation screening ECG and subsequent echocardiographic data was collected from 13,954 males aged 16 to 22, who reported for medical screening prior to military conscription. Final diagnosis of LVH was made on echocardiography, with LVH defined as a left ventricular mass index >115g/m2. The continuous and binary forms of classical criteria were compared against machine learning models using receiver-operating characteristics (ROC) curve analysis. An 80:20 split was used to divide the data into training and test sets for the machine learning models, and three fold cross validation was used in training the models. We also compared the important variables identified by machine learning models with the input variables of classical criteria. Results Prevalence of echocardiographic LVH in this population was 0.91% (127 cases). Classical ECG criteria had poor performance in predicting LVH, with the best predictions achieved by the continuous Sokolow-Lyon (AUC = 0.63, 95% CI = 0.58–0.68) and the continuous Modified Cornell (AUC = 0.63, 95% CI = 0.58–0.68). Machine learning methods achieved superior performance – Random Forest (AUC = 0.74, 95% CI = 0.66–0.82), Gradient Boosting Machines (AUC = 0.70, 95% CI = 0.61–0.79), GLMNet (AUC = 0.78, 95% CI = 0.70–0.86). Novel and less recognized ECG parameters identified by the machine learning models as being predictive of LVH included mean QT interval, mean QRS interval, R in V4, and R in I. ROC curves of models studies Conclusion The prevalence of LVH in our population is lower than that previously reported in other similar populations. Classical ECG criteria perform poorly in this context. Machine learning methods show superior predictive performance and demonstrate non-traditional predictors of LVH from ECG data. Further research is required to improve the predictive ability of machine learning models, and to understand the underlying pathology of the novel ECG predictors identified.

Sign in / Sign up

Export Citation Format

Share Document