At present, the majority of machine learning jobs involve working with large datasets. You can't do that using a single machine. So, you need to distribute across a cluster. Get acquainted with tools like Apache Hadoop, and cloud services like Rackspace, Amazon EC2, Google Cloud Platform, OpenStack, and Microsoft Azure etc. You should also master all of the great Unix tools such as cat, grep, find, awk, sed, sort, cut, tr etc. Since all of the processing will most likely be the on the Linux-based machine, you need access to learn these tools, their functions, and applications.

It is conventional wisdom in machine learning and data mining that logical models such as rule sets are more interpretable than other models, and that among such rule-based models, simpler models are more interpretable than more complex ones. In this position paper, we question this latter assumption by focusing on one particular aspect of interpretability, namely the plausibility of models. Roughly speaking, we equate the plausibility of a model with the likeliness that a user accepts it as an explanation for a prediction. In particular, we argue that, all other things being equal, longer explanations may be more convincing than shorter ones, and that the predominant bias for shorter models, which is typically necessary for learning powerful discriminative models, may not be suitable when it comes to user acceptance of the learned models. To that end, we first recapitulate evidence for and against this postulate, and then report the results of an evaluation in a crowd-sourcing study based on about 3.000 judgments. The results do not reveal a strong preference for simple rules, whereas we can observe a weak preference for longer rules in some domains. We then relate these results to well-known cognitive biases such as the conjunction fallacy, the representative heuristic, or the recogition heuristic, and investigate their relation to rule length and plausibility.

In scientific inference problems, the underlying statistical modeling assumptions have a crucial impact on the end results. There exist, however, only a few automatic means for validating these fundamental modelling assumptions. The contribution in this paper is a general criterion to evaluate the consistency of a set of statistical models with respect to observed data. This is achieved by automatically gauging the models' ability to generate data that is similar to the observed data. Importantly, the criterion follows from the model class itself and is therefore directly applicable to a broad range of inference problems with varying data types. The proposed data consistency criterion is illustrated and evaluated using three synthetic and two real data sets.

Hence, it is seldom the case where a single oracle labels an entire collection. Furthermore, the Web, through its social nature, also exploits the wisdom of crowds to annotate large collections of documents and images. By categorizing texts, tagging images or rating products and places, Web users are generating large volumes of labeled content. However, when learning supervised models from crowds, the quality of labels can vary significantly due to task subjectivity and differences in annotator reliability (or bias) [9], [10]. If we consider a sentiment analysis task, it becomes clear that the subjectiveness of the exercise is prone to generate considerably distinct labels from different annotators. Similarly, online product reviews are known to vary considerably depending on the personal biases and volatility of the reviewer's opinions. It is therefore essential to account for these issues when learning from this increasingly common type of data. Hence, the interest of researchers on building models that take the reliabilities of different annotators into consideration and mitigate the effect of their biases has spiked during the last few years (e.g.

Checking various log files from different processes can be a tedious task as these logs contain lots of events, each with a (possibly large) number of attributes. We developed a way to automatically model log files with a dozen attributes and detect outlier traces in the data. For that we extend Dynamic Bayesian Networks to model the normal behavior found in log files. We introduce a new algorithm that is able to learn a model of a log file starting from the data itself. The model is capable of scoring traces even when new values or new combinations of values appear in the log file and has the ability to give a decomposition of the score indicating the root cause for the anomalies.

Finally, a comprehensive hands-on machine learning course with specific focus on classification based models for the investment community and passionate investors. In the past few years, there has been a massive adoption and growth in the use of data science, artificial intelligence and machine learning to find alpha. However, information on and application of machine learning to investment are scarce. This course has been designed to address that. It is meant to spark your creative juices and get you started in this space.

Frequentist Statistics tests whether an event (hypothesis) occurs or not. It calculates the probability of an event in the long run of the experiment. A very common flaw found in frequentist approach i.e. dependence of the result of an experiment on the number of times the experiment is repeated. Bayesian statistics is a mathematical procedure that applies probabilities to statistical problems. It provides people the tools to update their beliefs in the evidence of new data.

The Dynamic Chain Event Graph (DCEG) is able to depict many classes of discrete random processes exhibiting asymmetries in their developments and context-specific conditional probabilities structures. However, paradoxically, this very generality has so far frustrated its wide application. So in this paper we develop an object-oriented method to fully analyse a particularly useful and feasibly implementable new subclass of these graphical models called the N Time-Slice DCEG (NT-DCEG). After demonstrating a close relationship between an NT-DCEG and a specific class of Markov processes, we discuss how graphical modellers can exploit this connection to gain a deep understanding of their processes. We also show how to read from the topology of this graph context-specific independence statements that can then be checked by domain experts. Our methods are illustrated throughout using examples of dynamic multivariate processes describing inmate radicalisation in a prison.

This paper studies the problem of {\em learning} the probability distribution $P_X$ of a discrete random variable $X$ using indirect and sequential samples. At each time step, we choose one of the possible $K$ functions, $g_1, \ldots, g_K$ and observe the corresponding sample $g_i(X)$. The goal is to estimate the probability distribution of $X$ by using a minimum number of such sequential samples. This problem has several real-world applications including inference under non-precise information and privacy-preserving statistical estimation. We establish necessary and sufficient conditions on the functions $g_1, \ldots, g_K$ under which asymptotically consistent estimation is possible. We also derive lower bounds on the estimation error as a function of total samples and show that it is order-wise achievable. Leveraging these results, we propose an iterative algorithm that i) chooses the function to observe at each step based on past observations; and ii) combines the obtained samples to estimate $p_X$. The performance of this algorithm is investigated numerically under various scenarios, and shown to outperform baseline approaches.

Dey et al. (2017), bayesian formulation of this problem is discussed. Hyper prameter choice was the major issue in that paper. However this problem was attempted only after suitable choice of feature dimension for user and movie feature vector. Usual method used to select such feature dimension is nothing but one dimensional grid search that select the dimension which minimizes the test error. This approach is boring as it attracts extra computational burden to select the feature dimension.