For machine learning, you can also use these libraries to build learning models. However, doing so requires that you have a strong appreciation of the mathematical foundation for the various machine learning algorithms.

Suppose I have a classification neuralnet for which I compute the confusion matrix on the validation set after my network has converged. What ways are there of using this matrix to reliably increase the accuracy on unseen data? I know of setting a per-class minimum confidence threshold. But would it make sense to reponder the softmax predictions knowing that some class A is often misclassified as B by the network etc...?

The presence of noisy instances in mobile phone data is a fundamental issue for classifying user phone call behavior (i.e., accept, reject, missed and outgoing), with many potential negative consequences. The classification accuracy may decrease and the complexity of the classifiers may increase due to the number of redundant training samples. To detect such noisy instances from a training dataset, researchers use naive Bayes classifier (NBC) as it identifies misclassified instances by taking into account independence assumption and conditional probabilities of the attributes. However, some of these misclassified instances might indicate usages behavioral patterns of individual mobile phone users. Existing naive Bayes classifier based noise detection techniques have not considered this issue and, thus, are lacking in classification accuracy. In this paper, we propose an improved noise detection technique based on naive Bayes classifier for effectively classifying users' phone call behaviors. In order to improve the classification accuracy, we effectively identify noisy instances from the training dataset by analyzing the behavioral patterns of individuals. We dynamically determine a noise threshold according to individual's unique behavioral patterns by using both the naive Bayes classifier and Laplace estimator. We use this noise threshold to identify noisy instances. To measure the effectiveness of our technique in classifying user phone call behavior, we employ the most popular classification algorithm (e.g., decision tree). Experimental results on the real phone call log dataset show that our proposed technique more accurately identifies the noisy instances from the training datasets that leads to better classification accuracy.

We present the design and implementation of a custom discrete optimization technique for building rule lists over a categorical feature space. Our algorithm produces rule lists with optimal training performance, according to the regularized empirical risk, with a certificate of optimality. By leveraging algorithmic bounds, efficient data structures, and computational reuse, we achieve several orders of magnitude speedup in time and a massive reduction of memory consumption. We demonstrate that our approach produces optimal rule lists on practical problems in seconds. Our results indicate that it is possible to construct optimal sparse rule lists that are approximately as accurate as the COMPAS proprietary risk prediction tool on data from Broward County, Florida, but that are completely interpretable. This framework is a novel alternative to CART and other decision tree methods for interpretable modeling.

We study fairness in collaborative-filtering recommender systems, which are sensitive to discrimination that exists in historical data. Biased data can lead collaborative-filtering methods to make unfair predictions for users from minority groups. We identify the insufficiency of existing fairness metrics and propose four new metrics that address different forms of unfairness. These fairness metrics can be optimized by adding fairness terms to the learning objective. Experiments on synthetic and real data show that our new metrics can better measure fairness than the baseline, and that the fairness objectives effectively help reduce unfairness.

While competing in a Kaggle competition this summer, I came across a simple visualization (created by a fellow competitor) that helped me to gain a better intuitive understanding of ROC curves and Area Under the Curve (AUC). I created a video explaining this visualization to serve as a learning aid for my Data Science students, and decided to share it publicly to help others understand this complex topic. An ROC curve is the most commonly used way to visualize the performance of a binary classifier, and AUC is (arguably) the best way to summarize its performance in a single number. As such, gaining a deep understanding of ROC curves and AUC is beneficial for data scientists, machine learning practitioners, and medical researchers (among others). The 14-minute video is embedded below, followed by the complete transcript (including graphics).

We propose stochastic, non-parametric activation functions that are fully learnable and individual to each neuron. Complexity and the risk of overfitting are controlled by placing a Gaussian process prior over these functions. The result is the Gaussian process neuron, a probabilistic unit that can be used as the basic building block for probabilistic graphical models that resemble the structure of neural networks. The proposed model can intrinsically handle uncertainties in its inputs and self-estimate the confidence of its predictions. Using variational Bayesian inference and the central limit theorem, a fully deterministic loss function is derived, allowing it to be trained as efficiently as a conventional neural network using mini-batch gradient descent. The posterior distribution of activation functions is inferred from the training data alongside the weights of the network. The proposed model favorably compares to deep Gaussian processes, both in model complexity and efficiency of inference. It can be directly applied to recurrent or convolutional network structures, allowing its use in audio and image processing tasks. As an preliminary empirical evaluation we present experiments on regression and classification tasks, in which our model achieves performance comparable to or better than a Dropout regularized neural network with a fixed activation function. Experiments are ongoing and results will be added as they become available.

Learning from many real-world datasets is limited by a problem called the class imbalance problem. A dataset is imbalanced when one class (the majority class) has significantly more samples than the other class (the minority class). Such datasets cause typical machine learning algorithms to perform poorly on the classification task. To overcome this issue, this paper proposes a new approach Neighbors Progressive Competition (NPC) for classification of imbalanced datasets. Whilst the proposed algorithm is inspired by weighted k-Nearest Neighbor (k-NN) algorithms, it has major differences from them. Unlike k- NN, NPC does not limit its decision criteria to a preset number of nearest neighbors. In contrast, NPC considers progressively more neighbors of the query sample in its decision making until the sum of grades for one class is much higher than the other classes. Furthermore, NPC uses a novel method for grading the training samples to compensate for the imbalance issue. The grades are calculated using both local and global information. In brief, the contribution of this paper is an entirely new classifier for handling the imbalance issue effectively without any manually-set parameters or any need for expert knowledge. Experimental results compare the proposed approach with five representative algorithms applied to fifteen imbalanced datasets and illustrate this algorithms effectiveness.

The metrics that you choose to evaluate your machine learning algorithms are very important. Choice of metrics influences how the performance of machine learning algorithms is measured and compared. They influence how you weight the importance of different characteristics in the results and your ultimate choice of which algorithm to choose. In this post, you will discover how to select and use different machine learning performance metrics in Python with scikit-learn. Metrics To Evaluate Machine Learning Algorithms in Python Photo by Ferrous Büller, some rights reserved.

Customer churn is a problem that all companies need to monitor, especially those that depend on subscription-based revenue streams. The simple fact is that most organizations have data that can be used to target these individuals and to understand the key drivers of churn, and we now have Keras for Deep Learning available in R (Yes, in R!!), which predicted customer churn with 82% accuracy. We're super excited for this article because we are using the new keras package to produce an Artificial Neural Network (ANN) model on the IBM Watson Telco Customer Churn Data Set! As for most business problems, it's equally important to explain what features drive the model, which is why we'll use the lime package for explainability. In addition, we use three new packages to assist with Machine Learning (ML): recipes for preprocessing, rsample for sampling data and yardstick for model metrics. These are relatively new additions to CRAN developed by Max Kuhn at RStudio (creator of the caret package). It seems that R is quickly developing ML tools that rival Python. Good news if you're interested in applying Deep Learning in R! We are so let's get going!! Customer churn refers to the situation when a customer ends their relationship with a company, and it's a costly problem. Customers are the fuel that powers a business. Further, it's much more difficult and costly to gain new customers than it is to retain existing customers. As a result, organizations need to focus on reducing customer churn. The good news is that machine learning can help. For many businesses that offer subscription based services, it's critical to both predict customer churn and explain what features relate to customer churn.