Success is a process not an event. Data Science is growing rapidly in all sectors. With the availability of so many technologies within the Data Science domain, it becomes tricky to crack any Data Science interview. In this article, we have tried to cover the most common Data Science interview questions asked by recruiters. Answer: The question can also be phrased as to why linear regression is not a very effective algorithm.

Batch Gradient Descent, Mini-batch Gradient Descent and Stochastic Gradient Descent are techniques used for gradient optimization differ in the batch size they use for computing gradients in each iteration. Gradient Descent uses all the data to compute gradients and update weights in each iteration. Minibatch Gradient Descent takes a subset of dataset to update its weights in each iteration. It however takes more iterations to converge to minima, but it is faster as compared to Gradient Descent due to lesser size of batch data used. Stochastic Gradient Descent (SGD) (or also sometimes on-line gradient descent) is the extreme case of this.

Sparse Bayesian Learning (SBL) constructs an extremely sparse probabilistic model with very competitive generalization. However, SBL needs to invert a big covariance matrix with complexity O(M^3 ) (M: feature size) for updating the regularization priors, making it difficult for practical use. There are three issues in SBL: 1) Inverting the covariance matrix may obtain singular solutions in some cases, which hinders SBL from convergence; 2) Poor scalability to problems with high dimensional feature space or large data size; 3) SBL easily suffers from memory overflow for large-scale data. This paper addresses these issues with a newly proposed diagonal Quasi-Newton (DQN) method for SBL called DQN-SBL where the inversion of big covariance matrix is ignored so that the complexity and memory storage are reduced to O(M). The DQN-SBL is thoroughly evaluated on non-linear classifiers and linear feature selection using various benchmark datasets of different sizes. Experimental results verify that DQN-SBL receives competitive generalization with a very sparse model and scales well to large-scale problems.

Deep Gaussian Processes (DGPs) are multi-layer, flexible extensions of Gaussian processes but their training remains challenging. Sparse approximations simplify the training but often require optimization over a large number of inducing inputs and their locations across layers. In this paper, we simplify the training by setting the locations to a fixed subset of data and sampling the inducing inputs from a variational distribution. This reduces the trainable parameters and computation cost without significant performance degradations, as demonstrated by our empirical results on regression problems. Our modifications simplify and stabilize DGP training while making it amenable to sampling schemes for setting the inducing inputs.

Probabilistic graphical models are a fundamental tool in probabilistic modeling, machine learning and artificial intelligence. They allow us to integrate in a natural way expert knowledge, physical modeling, heterogeneous and correlated data and quantities of interest. For exactly this reason, multiple sources of model uncertainty are inherent within the modular structure of the graphical model. In this paper we develop information-theoretic, robust uncertainty quantification methods and non-parametric stress tests for directed graphical models to assess the effect and the propagation through the graph of multi-sourced model uncertainties to quantities of interest. These methods allow us to rank the different sources of uncertainty and correct the graphical model by targeting its most impactful components with respect to the quantities of interest. Thus, from a machine learning perspective, we provide a mathematically rigorous approach to correctability that guarantees a systematic selection for improvement of components of a graphical model while controlling potential new errors created in the process in other parts of the model. We demonstrate our methods in two physico-chemical examples, namely quantum scale-informed chemical kinetics and materials screening to improve the efficiency of fuel cells.

We study the problem of post-hoc calibration for multiclass classification, with an emphasis on histogram binning. Multiple works have focused on calibration with respect to the confidence of just the predicted class (or 'top-label'). We find that the popular notion of confidence calibration [Guo et al., 2017] is not sufficiently strong -- there exist predictors that are not calibrated in any meaningful way but are perfectly confidence calibrated. We propose a closely related (but subtly different) notion, top-label calibration, that accurately captures the intuition and simplicity of confidence calibration, but addresses its drawbacks. We formalize a histogram binning (HB) algorithm that reduces top-label multiclass calibration to the binary case, prove that it has clean theoretical guarantees without distributional assumptions, and perform a methodical study of its practical performance. Some prediction tasks require stricter notions of multiclass calibration such as class-wise or canonical calibration. We formalize appropriate HB algorithms corresponding to each of these goals. In experiments with deep neural nets, we find that our principled versions of HB are often better than temperature scaling, for both top-label and class-wise calibration. Code for this work will be made publicly available at https://github.com/aigen/df-posthoc-calibration.

Federated learning (FL) is a distributed model for deep learning that integrates client-server architecture, edge computing, and real-time intelligence. FL has the capability of revolutionizing machine learning (ML) but lacks in the practicality of implementation due to technological limitations, communication overhead, non-IID (independent and identically distributed) data, and privacy concerns. Training a ML model over heterogeneous non-IID data highly degrades the convergence rate and performance. The existing traditional and clustered FL algorithms exhibit two main limitations, including inefficient client training and static hyper-parameter utilization. To overcome these limitations, we propose a novel hybrid algorithm, namely genetic clustered FL (Genetic CFL), that clusters edge devices based on the training hyper-parameters and genetically modifies the parameters cluster-wise. Then, we introduce an algorithm that drastically increases the individual cluster accuracy by integrating the density-based clustering and genetic hyper-parameter optimization. The results are bench-marked using MNIST handwritten digit dataset and the CIFAR-10 dataset. The proposed genetic CFL shows significant improvements and works well with realistic cases of non-IID and ambiguous data.

We can make use of Machine Learning in order to categorize such harassment incidents without any manual human intervention. Traditionally if a report needs to file regarding a harassment case, the victim needs to provide the description of the case along with filling up several forms. We can automate some parts of this process by categorizing the incidents based on the descriptions and then performing further processes based on the category of the cases. This automatic classification of different forms of sexual harassment will immensely help the authorities as well as the concerned organizations to partially automate their systems that file reports against such incidents. Here our objective is to use short stories which were submitted online and to be able to automatically categorize each story submitted by a user.

In the context of classification problems, a softmax classifier with a cross-entropy loss is often the go-to approach. However, in situations with many classes, softmax can be slow to train as it requires an output node for every class, leading to very large output layers. For example, a network with a hidden layer size of 300 and 100,000 output classes has 30 million parameters in the output layer alone. In applied AI settings, these types of problems occur often. An example of this is learning to match papers with authors, or product descriptions with actual products.

Introduction Since we productionized distributed XGBoost on Apache Spark™ at Uber in 2017, XGBoost has powered a wide spectrum of machine learning (ML) use cases at Uber, spanning from optimizing marketplace dynamic pricing policies for Freight, improving times of arrival (ETA) estimation, fraud detection and prevention, to content discovery and recommendation for Uber Eats. However, as Uber has scaled, we have started to run distributed training jobs with more data and workers, and more complex distributed training patterns have become increasingly common. As such, we have observed a number of challenges for doing distributed machine learning and deep learning at