DOCSLIB.ORG

Topics in Representation Learning

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Topics in Representation Learning The Raymond and Beverly Sackler Faculty of Exact Sciences The Blavatnik School of Computer Science Topics in Representation Learning Thesis submitted for the degree of “Doctor of Philosophy” by Oren Barkan This thesis was carried out under the supervision of Prof. Amir Averbuch Submitted to the Senate of Tel-Aviv University October, 2018 Abstract The success of machine learning algorithms depends on the input representation. Over the last few decades, a large amount of effort has been spent by humankind in finding and designing informative features. These handcrafted features are task dependent and require different types of expertise in different domains. Representation learning algorithms aim at discovering useful informative features, automatically. This is done by learning linear and nonlinear transformations from the high dimensional input feature space to a lower dimensional latent space, in which the important information in the data is preserved. This thesis presents several developments in representation learning for various applications: face recognition, out-of-sample extension, word embedding and item recommendations. In the first part, we advance descriptor based face recognition methods by several aspects: first, we propose a novel usage of over-complete representations and show that it improves classification accuracy when coupled with dimensionality reduction. Second, we introduce a new supervised metric learning pipeline. This is further extended to the unsupervised case, where the data lacks labels. Finally, we propose the application of the manifold learning algorithm, Diffusion Maps, as a nonlinear alternative to Whitened Principal Component Analysis, which is a common practice in face recognition systems. We demonstrate the effectiveness of the proposed method on real world data. The second part of the thesis proposes a Bayesian nonparametric method for the out- of-sample problem, which is a common problem that exists in manifold learning algorithms. The method is based on Gaussian Process Regression, naturally provides a measure of confidence in the prediction and is independent of the manifold learning algorithm. The connection between our method and the commonly used Nystrom extension is analyzed, where it is shown that the latter is a special case of the former. We validate our method in a series of experiments on synthetic and real world datasets. In the third part, we propose a Bayesian neural word embedding algorithm. The algorithm is based on a Variational Bayes solution to the Skip-Gram objective that is commonly used for mapping words to vectors (word2vec) in a latent vector space. In this space, semantic and syntactic relations between words can be inferred via inner product between word vectors. Different from Skip-Gram, our method maps words to probability density functions in a latent space. Furthermore, our method is scalable and enables parameter updates that are embarrassingly parallel. We evaluate our method on words analogy and similarity tasks and show that it is competitive with the original Skip-Gram method. In the fourth part, we propose a multiview neural item embedding method for mapping multiple content based representations of items to their collaborative filtering representation. We introduce a multiview deep regression model that is capable of learning item similarities from a mix of categorical, continuous and unstructured data. Our model is trained to map these content based information sources to a latent space that is induced by collaborative filtering relations. We demonstrate the effectiveness of the proposed model in predicting useful item recommendations and similarities and explore the contribution of each of its components and their combinations. Furthermore, we show that our model outperforms a model that is based on content solely. Acknowledgements I want to thank all those who helped and supported me during my PhD studies. First and foremost, I thank my advisor, Amir Averbuch for the guidance, mentoring and support. You have always had great intuition and insights on algorithmic problems and open mind regarding various research fields. You taught me how to ask the right questions and conduct proper research that eventually leads to elegant solutions. I want to thank my research collaborators during the years: Noam Koenigstein, Lior Wolf, Shai Dekel, Jonathan Weill, Hagai Aronowitz, Eylon Yogev, Nir Nice, Yael Brumer, David Tsiris, Shay Ben-Elazar, Ori Kats and Amjad Abu-Rmileh. It was a pleasure working with you and I hope we will keep collaborating in the future. Finally, I would like to thank my parents and the rest of the family for their endless support and encouragement. To my grandfather Contents Introduction 1 Outline and Contributions of this Thesis 6 Published and Submitted Papers 10 Funding Acknowledgements 11 1 Learning Latent Face Representations in High Dimensional Feature Spaces 12 1.1 Introduction and Related Work 12 1.1.1 Outline of modern face verification systems 12 1.1.2 Fisher’s Linear Discriminant Analysis 16 1.1.3 Labeled Faces in the Wild (LFW) 17 1.1.4 Outline and contributions of this chapter 18 1.2 Over-complete Representations 20 1.2.1 Outline of modern face verification systems 20 1.2.2 Outline of modern face verification systems 23 1.3 Within Class Covariance Normalization (WCCN) 23 1.4 Leveraging Unlabeled Data for Supervised Classification 24 1.5 The Proposed Recognition Pipeline 26 1.5.1 The unsupervised pipeline 27 1.5.2 Manifold learning in the descriptor space via Diffusion Maps 28 1.5.2.1 Out of sample extension 30 1.6 Experimental Setup and Results 31 1.6.1 Front-end processing 32 1.6.2 The evaluated descriptors 32 1.6.3 System parameters 32 1.6.4 Results 33 1.7 Conclusion 39 2 Gaussian Process Regression for Out-of-Sample Extension 41 2.1 Introduction 41 2.2 Related Work 43 2.3 Gaussian Process Regression (GPR) 44 2.4 Gaussian Process Regression based Out-of-Sample Extension 45 2.4.1 The connection between GPR and Nystrom extension 46 2.5 Experimental Setup and Results 49 2.5.1 The experimental workflow 49 2.5.2 The evaluated OOSE methods 49 2.5.3 The evaluated manifold learning algorithms 50 2.5.4 Datasets 50 2.5.5 Experiment 1 50 2.5.6 Experiment 2 51 2.5.7 Experiment 3 52 2.5.8 Experiment 4 52 2.6 Conclusion 55 3 Bayesian Neural Word Embedding 57 3.1 Introduction and Related Work 57 3.2 Skip-Gram with Negative Sampling 58 3.2.1 Negative sampling 59 3.2.2 Data subsampling 60 3.2.3 Word representation and similarity 60 3.3 Bayesian Skip-Gram (BSG) 61 3.3.1 Variational approximation 62 3.3.2 Stochastic updates 65 3.4 The BSG Algorithm 65 3.4.1 Stage 1 - initialization 67 3.4.2 Stage 2 - data sampling 67 3.4.3 Stage 3 - parameter updates 68 3.4.4 Similarity measures 68 3.5 Experimental Setup and Results 70 3.5.1 Datasets 71 3.5.2 Parameter configuration 71 3.5.3 Results 71 3.6 Conclusion 73 4 Multiview Neural Item Embedding 74 4.1 Introduction 74 4.2 Related Work 78 4.3 Item2vec - SGNS for item based CF 80 4.4 CB2CF - Deep Multiview Regression Model 83 4.4.1 Text components 84 4.4.1 Tags components 86 4.4.1 Numeric components 87 4.4.1 The combiner components 87 4.4.1 The full model (CB2CF) 87 4.5 Experimental Setup and Results 88 4.5.1 Datasets 89 4.5.1.1 Word2vec dataset 89 4.5.1.2 Movies dataset 89 4.5.1.3 Windows apps dataset 90 4.5.2 Evaluated systems 90 4.5.3 Parameter configuration 91 4.5.4 Evaluation measures 91 4.5.5 Quantitative results - Experiment 1 93 4.5.6 Quantitative results - Experiment 2 95 4.5.7 Qualitative results 98 4.6 Conclusion 102 4.7 Future Research 102 Conclusions 104 Bibliography 107 Introduction The performance of machine learning algorithms heavily depends on the input data representation. Often the raw data is not in the form that enables a straightforward application of learning algorithms. As a result, a large amount of effort is spent in order to bring the data to a useful form. The process of feature engineering is domain- dependent, where different types of data require different types of expertise. Therefore, any method that is capable of extracting ‘good’ features, automatically, is considered to be valuable. While in many applications the acquired data is of high dimensionality, the intrinsic information which resides in the measurements is of much lower dimensionality [50]- [53]. Indeed, in many types of real world data, most of the variability can be captured by a low dimensional manifold and in some scenarios, even by a linear subspace [120]- [129]. Representation learning algorithms attempt to discover the intrinsic properties of the data. This is done by learning informative features that preserve the intrinsic geometry of the data, while discarding a large amount of redundancy [130]. The result is an organized low dimensional representation which enables a better understanding and efficient inspection of the data. During the last two decades, a major breakthrough has revolutionized the field of artificial intelligence. The development of novel representation learning algorithms significantly advanced the state-of-the-art in various domains such as computer vision [65, 41, 77, 134], speech [70, 78, 120, 121], natural language processing [71, 72, 74, 107], recommender systems [66, 69, 80, 111] and manifold learning [50]-[53]. 1 This thesis proposes models for several different tasks: in the domain of computer vision, we propose models for unconstrained face verification. In the domain of manifold learning, we propose a general Bayesian model for out-of-sample extension.
Load more

Recommended publications
  • Advanced Topics in Learning and Vision
    Advanced Topics in Learning and Vision Ming-Hsuan Yang [email protected] Lecture 6 (draft) Announcements • More course material available on the course web page • Project web pages: Everyone needs to set up one (format details will soon be available) • Reading (due Nov 1): - RVM application for 3D human pose estimation [1]. - Viola and Jones: Adaboost-based real-time face detector [25]. - Viola et al: Adaboost-based real-time pedestrian detector [26]. A. Agarwal and B. Triggs. 3D human pose from silhouettes by relevance vector regression. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, volume 2, pages 882–888, 2004. P. Viola and M. Jones. Robust real-time face detection. International Journal of Computer Vision, 57(2):137–154, 2004. P. Viola, M. Jones, and D. Snow. Markov face models. In Proceedings of the Ninth IEEE International Conference on Computer Vision, volume 2, pages 734–741, 2003. Lecture 6 (draft) 1 Overview • Linear classifier: Fisher linear discriminant (FLD), linear support vector machine (SVM). • Nonlinear classifier: nonlinear support vector machine, • SVM regression, relevance vector machine (RVM). • Kernel methods: kernel principal component, kernel discriminant analysis. • Ensemble of classifiers: Adaboost, bagging, ensemble of homogeneous/heterogeneous classifiers. Lecture 6 (draft) 2 Fisher Linear Discriminant • Based on Gaussian distribution: between-scatter/within-scatter matrices. • Supervised learning for multiple classes. 0 • Trick often used in ridge regression: Sw = Sw + λI without using PCA • Fisherfaces vs. Eigenfaces (FLD vs. PCA). • Further study: Aleix Martinez’s paper on analyzing FLD for object recognition [10][11]. A. M. Martinez and A. Kak. PCA versus LDA.
    [Show full text]
  • Dropout-Based Automatic Relevance Determination
    Dropout-based Automatic Relevance Determination Dmitry Molchanov1, Arseniy Ashuha2 and Dmitry Vetrov3;4∗ 1Skolkovo Institute of Science and Technology 2Moscow Institute of Physics and Technology 3National Research University Higher School of Economics, 4Yandex Moscow, Russia [email protected], [email protected], [email protected] 1 Introduction During past several years, several works about treating dropout as a Bayesian model have appeared [1,3,4]. Authors of [1] proposed a method that represents dropout as a special case of Bayesian model and allows to tune individual dropout rates for each weight. Another direction of research is the exploration of the ways to train neural networks with sparse weight matrices [6,9] and the ways to prune excessive neurons [8]. One of the possible approaches is so-called Automatic Relevance Determination (ARD). A typical example of such models is the Relevance Vector Machine [4]. However, these models are usually relatively hard to train. In this work, we introduce a new dropout-based way to obtain the ARD effect that can be applied to complex models via stochastic variational inference framework. 2 Theory Variational dropout [1] is a recent technique that generalizes Gaussian dropout [7]. It allows to set individual dropout rates for each weight/neuron/layer and to tune them with a simple gradient descent based method. As in most modern Bayesian models, the goal is to obtain the variational approximation to the posterior distribution via optimization of the variational lower bound (1): L(q) = q(w) log p(T j X; w) − DKL(q(w) k pprior(w)) ! max (1) E q To put it simply, the main idea of Variational dropout is to search for posterior approximation in a 2 specific family of distributions: q(wi) = N (θi; αiθi ).
    [Show full text]
  • Healing the Relevance Vector Machine Through Augmentation
    Healing the Relevance Vector Machine through Augmentation Carl Edward Rasmussen [email protected] Max Planck Institute for Biological Cybernetics, 72076 T¨ubingen, Germany Joaquin Qui˜nonero-Candela [email protected] Friedrich Miescher Laboratory, Max Planck Society, 72076 T¨ubingen, Germany Abstract minimizing KL divergences between the approximated and true posterior, by Smola and Sch¨olkopf (2000) and The Relevance Vector Machine (RVM) is a Smola and Bartlett (2001) by making low rank approx- sparse approximate Bayesian kernel method. imations to the posterior. In (Gibbs & MacKay, 1997) It provides full predictive distributions for and in (Williams & Seeger, 2001) sparseness arises test cases. However, the predictive uncer- from matrix approximations, and in (Tresp, 2000) tainties have the unintuitive property, that from neglecting correlations. The use of subsets of they get smaller the further you move away regressors has also been suggested by Wahba et al. from the training cases. We give a thorough (1999). analysis. Inspired by the analogy to non- degenerate Gaussian Processes, we suggest The Relevance Vector Machine (RVM) introduced by augmentation to solve the problem. The pur- Tipping (2001) produces sparse solutions using an im- pose of the resulting model, RVM*, is primar- proper hierarchical prior and optimizing over hyperpa- ily to corroborate the theoretical and exper- rameters. The RVM is exactly equivalent to a Gaus- imental analysis. Although RVM* could be sian Process, where the RVM hyperparameters are pa- used in practical applications, it is no longer rameters of the GP covariance function (more on this a truly sparse model. Experiments show that in the discussion section).
    [Show full text]
  • Sparse Bayesian Classification with the Relevance Vector Machine
    Sparse Bayesian Classification with the Relevance Vector Machine In a pattern recognition setting, the Relevance Vector Machine (RVM) [3] can be viewed as a simple logistic regression model, with a Bayesian Automatic Relevance Determination (ARD) prior [1] over the weights associated with each feature in order to achieve a parsimonious model. Let A = {A, C, G, T} represent an alphabet of symbols representing the bases adenine, cytosine, guanine and thymine respectively. The RVM constructs a decision rule from a set of labelled training data, ` ni D = {(xi, ti)}i=1, xi ∈ A , ti ∈ {0, + 1}, where the input patterns, xi, consist of strings drawn from A of varying lengths, describing the upstream promoter regions of a set of co-regulated genes. The target patterns, ti, indicate whether the corresponding gene is up-regulated (class C1, yi = +1) or down-regulated (class C2, yi = 0) under a given treatment. The RVM constructs a logistic regression model based on a set of sequence features derived from the input patterns, i.e. N X p(C1|x) ≈ σ {y(x; w)} where y(x; w) = wiϕi(x) + w0, (1) i=1 and σ {y} = (1 + exp{y})−1 is the logistic inverse link function. In this study a d feature, ϕi(x), represents the number of times an arbitrary substring, si ∈ A , ocurrs in a promoter sequence x. A sufficiently large set of features is used such that it is reasonable to expect that some of these features will represent oligonucleotides forming a relevant promoter protein binding site and so provide discriminatory information for the pattern recognition task at hand.
    [Show full text]
  • Visual Sparse Bayesian Reinforcement Learning: a Framework for Interpreting What an Agent Has Learned
    Visual Sparse Bayesian Reinforcement Learning: A Framework for Interpreting What an Agent Has Learned Indrajeet Mishra, Giang Dao, and Minwoo Lee Department of Computer Science University of North Carolina at Charlotte {imishra, gdao, minwoo.lee}@uncc.edu Abstract— This paper presents a Visual Sparse Bayesian problem of deep learning as lack of visualization tools to un- Reinforcement Learning (V-SBRL) framework for recording derstand the computation performed in the hidden layers. For the images of the most important memories from the past this reason, many recent researches [10], [11], [12], [13], [14] experience. The key idea of this paper is to maintain an image snapshot storage to help understanding and analyzing have attempted different types of visualization to explore the the learned policy. In the extended framework of SBRL [1], learning process of a deep learning model. However, these the agent perceives the environment as the image state inputs, methods focus on visualizing how the features are computed encodes the image into feature vectors, train SBRL module in the intermediate layers in supervised learning, so it is and stores the raw images. In this process, the snapshot not enough to fully explain what it has learned and what storage keeps only the relevant memories which are important to make future decisions and discards the not-so-important the learned knowledge represents in reinforcement learning. memories. The stored snapshot images enable us to understand Although the recent visualization efforts for reinforcement the agent’s learning process by visualizing them. They also learning [15], [16] provides ways to interpret the learning of provide explanation of exploited policy in different conditions.
    [Show full text]
  • Modular Meta-Learning with Shrinkage
    Modular Meta-Learning with Shrinkage Yutian Chen∗ Abram L. Friesen∗ Feryal Behbahani Arnaud Doucet David Budden Matthew W. Hoffman Nando de Freitas DeepMind London, UK {yutianc, abef}@google.com Abstract Many real-world problems, including multi-speaker text-to-speech synthesis, can greatly benefit from the ability to meta-learn large models with only a few task- specific components. Updating only these task-specific modules then allows the model to be adapted to low-data tasks for as many steps as necessary without risking overfitting. Unfortunately, existing meta-learning methods either do not scale to long adaptation or else rely on handcrafted task-specific architectures. Here, we propose a meta-learning approach that obviates the need for this often sub-optimal hand-selection. In particular, we develop general techniques based on Bayesian shrinkage to automatically discover and learn both task-specific and general reusable modules. Empirically, we demonstrate that our method discovers a small set of meaningful task-specific modules and outperforms existing meta- learning approaches in domains like few-shot text-to-speech that have little task data and long adaptation horizons. We also show that existing meta-learning methods including MAML, iMAML, and Reptile emerge as special cases of our method. 1 Introduction The goal of meta-learning is to extract shared knowledge from a large set of training tasks to solve held-out tasks more efficiently. One avenue for achieving this is to learn task-agnostic modules and reuse or repurpose these for new tasks. Reusing or repurposing modules can reduce overfitting in low-data regimes, improve interpretability, and facilitate the deployment of large multi-task models on limited-resource devices as parameter sharing allows for significant savings in memory.
    [Show full text]
  • Sparse Bayesian Learning for Predicting Phenotypes and Identifying Influential Markers
    Sparse Bayesian Learning for Predicting Phenotypes and Identifying Influential Markers by Maryam Ayat A thesis submitted to The Faculty of Graduate Studies of The University of Manitoba in partial fulfillment of the requirements of the degree of Doctor of Philosophy Department of Computer Science The University of Manitoba Winnipeg, Manitoba, Canada December 2018 c Copyright 2018 by Maryam Ayat Thesis advisor Author Michael Domaratzki Maryam Ayat Sparse Bayesian Learning for Predicting Phenotypes and Identifying Influential Markers Abstract In bioinformatics, Genomic Selection (GS) and Genome-Wide Association Studies (GWASs) are two related problems that can be applied to the plant breeding industry. GS is a method to predict phenotypes (i.e., traits) such as yield and disease resis- tance in crops from high-density markers positioned throughout the genome of the varieties. By contrast, a GWAS involves identifying markers or genes that underlie the phenotypes of importance in breeding. The need to accelerate the development of improved varieties, and challenges such as discovering all sorts of genetic factors related to a trait, increasingly persuade researchers to apply state-of-the-art machine learning methods to GS and GWASs. The aim of this study is to employ sparse Bayesian learning as a technique for GS and GWAS. The sparse Bayesian learning uses Bayesian inference to obtain sparse solutions in regression or classification problems. This learning method is also called the Relevance Vector Machine (RVM), as it can be viewed as a kernel-based model of identical form to the renowned Support Vector Machine (SVM) method. ii The RVM has some advantages that the SVM lacks, such as having probabilistic outputs, providing a much sparser model, and the ability to work with arbitrary kernel functions.
    [Show full text]
  • Arxiv:1911.06028V2 [Cs.LG] 7 May 2021
    Published as a conference paper at ICLR 2021 1.2 1.2 1.2 1.6 1.6 1.0 1.0 1.0 1.4 1.4 1.2 1.2 0.8 0.8 0.8 1.0 1.0 0.6 0.6 0.6 0.8 0.8 0.4 0.4 0.4 0.6 0.6 0.4 0.4 0.2 0.2 0.2 0.2 0.2 0 0 0 0 0 -0.2 -0.2 -0.2 -1.2 -0.8 -0.6 -0.2 0.4 0.8 -0.2 -1.2 -0.8 -0.6 -0.2 0.4 0.8 -0.2 -1.2 -0.8 -0.6 -0.2 0.4 0.8 -1.5 -1.0 -0.5 0 0.5 1.0 -1.5 -1.0 -0.5 0 0.5 1.0 (a) FC + softmax (b) Gaussian process (c) RVM (d) Discriminative GMM (e) SDGM (ours) Figure 1: Comparison of decision boundaries. The black and green circles represent training samples from classes 1 and 2, respectively. The dashed black line indicates the decision boundary between classes 1 and 2 and thus satisfies P (c = 1 j x) = (c = 2 j x) = 0:5. The dashed blue and red lines represent the boundaries between the posterior probabilities of the mixture components. learning. This learning algorithm reduces memory usage without losing generalization capability by obtaining sparse weights while maintaining the multimodality of the mixture model. The technical highlight of this study is twofold: One is that the SDGM finds the multimodal structure in the feature space and the other is that redundant Gaussian components are removed owing to sparse learning.
    [Show full text]
  • A Novel Integrated Approach of Relevance Vector Machine Optimized by Imperialist Competitive Algorithm for Spatial Modeling of Shallow Landslides
    remote sensing Article A Novel Integrated Approach of Relevance Vector Machine Optimized by Imperialist Competitive Algorithm for Spatial Modeling of Shallow Landslides Dieu Tien Bui 1,2, Himan Shahabi 3,* , Ataollah Shirzadi 4 , Kamran Chapi 4, Nhat-Duc Hoang 5, Binh Thai Pham 6 , Quang-Thanh Bui 7, Chuyen-Trung Tran 8, Mahdi Panahi 9 , Baharin Bin Ahamd 10 and Lee Saro 11,12,* 1 Geographic Information Science Research Group, Ton Duc Thang University, Ho Chi Minh City, Vietnam; [email protected] 2 Faculty of Environment and Labour Safety, Ton Duc Thang University, Ho Chi Minh City, Vietnam 3 Department of Geomorphology, Faculty of Natural Resources, University of Kurdistan, Sanandaj 66177-15175, Iran 4 Department of Rangeland and Watershed Management, Faculty of Natural Resources, University of Kurdistan, Sanandaj 66177-15175, Iran; [email protected] (A.S.); [email protected] (K.C.) 5 Faculty of Civil Engineering, Institute of Research and Development, Duy Tan University, P809-K7/25 Quang Trung, Danang 550000, Vietnam; [email protected] 6 Geotechnical Engineering and Artificial Intelligence Research Group (GEOAI), University of Transport Technology, Hanoi 100803, Vietnam; [email protected] 7 Faculty of Geography, VNU University of Science, 334 Nguyen Trai, ThanhXuan, Hanoi 100803, Vietnam; [email protected] 8 Faculty of Information Technology, Hanoi University of Mining and Geology, Pho Vien, Bac Tu Liem, Hanoi 100803, Vietnam; [email protected] 9 Young Researchers and Elites Club, North Tehran Branch, Islamic Azad University,
    [Show full text]
  • Software Reliability Prediction Via Relevance Vector Regression
    Neurocomputing 186 (2016) 66–73 Contents lists available at ScienceDirect Neurocomputing journal homepage: www.elsevier.com/locate/neucom Software reliability prediction via relevance vector regression Jungang Lou a,b, Yunliang Jiang b,n, Qing Shen b, Zhangguo Shen b, Zhen Wang c, Ruiqin Wang b a Institute of Cyber-Systems and Control, Zhejiang Univeristy, 310027 Hangzhou, China b School of Information Engineering, Huzhou University, 313000 Huzhou, China c College of Computer Science and Technology, Shanghai University of Electric Power, 200090 Shanghai, China article info abstract Article history: The aim of software reliability prediction is to estimate future occurrences of software failures to aid in Received 21 September 2015 maintenance and replacement. Relevance vector machines (RVMs) are kernel-based learning methods Received in revised form that have been successfully adopted for regression problems. However, they have not been widely 27 November 2015 explored for use in reliability applications. This study employs a RVM-based model for software relia- Accepted 9 December 2015 bility prediction so as to capture the inner correlation between software failure time data and the nearest Communicated by Liang Wang Available online 6 January 2016 m failure time data. We present a comparative analysis in order to evaluate the RVMs effectiveness in forecasting time-to-failure for software products. In addition, we use the Mann-Kendall test method to Keywords: explore the trend of predictive accuracy as m varies. The reasonable value range of m is achieved through Software reliability model paired T-tests in 10 frequently used failure datasets from real software projects. Relevance vector machine & 2016 Elsevier B.V.
    [Show full text]
  • Automatic Machine Learning Derived from Scholarly Big Data
    Automatic Machine Learning Derived from Scholarly Big Data Asnat Greenstein-Messicaa,b,∗, Roman Vainshteina,b, Gilad Katza,b, Bracha Shapiraa,b, Lior Rokacha,b aSoftware and Information Engineering Department, Ben Gurion University of the Negev bBe’er Sheva, Israel Abstract One of the challenging aspects of applying machine learning is the need to identify the algorithms that will perform best for a given dataset. This process can be difficult, time consuming and often requires a great deal of domain knowledge. We present Sommelier, an expert system for recom- mending the machine learning algorithms that should be applied on a previously unseen dataset. Sommelier is based on word embedding representations of the domain knowledge extracted from a large corpus of academic publications. When presented with a new dataset and its problem descrip- tion, Sommelier leverages a recommendation model trained on the word embedding representation to provide a ranked list of the most relevant algorithms to be used on the dataset. We demonstrate Sommelier’s effectiveness by conducting an extensive evaluation on 121 publicly available datasets and 53 classification algorithms. The top algorithms recommended for each dataset by Sommelier were able to achieve on average 97.7% of the optimal accuracy of all surveyed algorithms. Keywords: scholarly big data, algorithm recommendation, word embedding, expert system 1. Introduction The enormous growth in the creation of digital data has created numerous opportunities. Com- panies and organizations are now capable of mining data regarding almost every aspect of their activity. However, the ability to collect data far outstrips the ability of organizations to apply analytics or derive meaningful insights.
    [Show full text]
  • An Integrated Transfer Learning and Multitask Learning Approach For
    An Integrated Transfer Learning and Multitask Learning Approach for Pharmacokinetic Parameter Prediction Zhuyifan Ye1, Yilong Yang1,2, Xiaoshan Li2, Dongsheng Cao3, Defang Ouyang1* 1State Key Laboratory of Quality Research in Chinese Medicine, Institute of Chinese Medical Sciences (ICMS), University of Macau, Macau, China 2Department of Computer and Information Science, Faculty of Science and Technology, University of Macau, Macau, China 3Xiangya School of Pharmaceutical Sciences, Central South University, No. 172, Tongzipo Road, Yuelu District, Changsha, People’s Republic of China Note: Zhuyifan Ye and Yilong Yang made equal contribution to the manuscript. Corresponding author: Defang Ouyang, email: [email protected]; telephone: 853-88224514. ABSTRACT: Background: Pharmacokinetic evaluation is one of the key processes in drug discovery and development. However, current absorption, distribution, metabolism, excretion prediction models still have limited accuracy. Aim: This study aims to construct an integrated transfer learning and multitask learning approach for developing quantitative structure-activity relationship models to predict four human pharmacokinetic parameters. Methods: A pharmacokinetic dataset included 1104 U.S. FDA approved small molecule drugs. The dataset included four human pharmacokinetic parameter subsets (oral bioavailability, plasma protein binding rate, apparent volume of distribution at steady-state and elimination half-life). The pre-trained model was trained on over 30 million bioactivity data. An integrated transfer learning and multitask learning approach was established to enhance the model generalization. Results: The pharmacokinetic dataset was split into three parts (60:20:20) for training, validation and test by the improved Maximum Dissimilarity algorithm with the representative initial set selection algorithm and the weighted distance function. The multitask learning techniques enhanced the model predictive ability.
    [Show full text]