Experiments were run with a 25-fold cross-validation.

This paper proposes a self-explainable Deep Learning (SE-DL) system for an image classification problem that performs self-error detection. The self-error detection is key to improving the DL system's safe operation, especially in safety-critical applications such as automotive systems. A SE-DL system outputs both the class prediction and an explanation for that prediction, which provides insight into how the system makes its predictions for humans. Additionally, we leverage the explanation of the proposed SE-DL system to detect potential class prediction errors of the system. The proposed SE-DL system uses a set of concepts to generate the explanation. The concepts are human-understandable lower-level image features in each input image relevant to the higher-level class of that image. We present a concept selection methodology for scoring all concepts and selecting a subset of them based on their contribution to the error detection performance of the proposed SE-DL system. Finally, we present different error detection schemes using the proposed SE-DL system to compare them against an error detection scheme without any SE-DL system.

In the first part of this series, we explored how to use pre-trained models to classify images. In this second part, we will build our own classifiers from scratch, in order to understand some of the underlying techniques used to constructing image classification models. When we are working on an image classification problem, it is important that we understand the problem and the data. So, before jumping in and building our model, let's first try to understand what we are trying to build and why. We will be using this animal dataset from Kaggle.

Classification is one of the most widely used methods of identifying trends in datasets. These types of algorithms work towards the objective of classifying datasets reliably without employing other data manipulation techniques. Classification algorithms are either Supervised or Unsupervised. Supervised learning uses labelled data to make decisions whereas the latter uses unlabelled data. For E.g., Supervised learning may contend with linear relationships whereas unsupervised will have to contend with unknown relationships (usually clustering). Therefore, depending on the type of dataset, the analyst will have to decide on the appropriate algorithm to employ for a problem.

The best way to learn is to do projects. Especially for self-learners who learned some programming and tools of data science and wondering what next. I suggest, start doing some projects of your own. Also, for a beginner who has no industry experience, some practice projects can make a good portfolio. But for a beginner, it may be hard to think of a good project idea.

A good deep learning model has a carefully carved architecture. It needs enormous training data, effective hardware, skilled developers, and a vast amount of time to train and hyper-tune the model to achieve satisfactory performance. Therefore, building a deep learning model from scratch and training is practically impossible for every deep learning task. Here comes the power of Transfer Learning. Transfer Learning is the approach of making use of an already trained model for a related task.

Hackathons are a great way to learn new skills and build your personal brand. They push you to put your creative hats on and apply your theoretical and practical knowledge in an innovative manner. Generally, the problems given at Hackathons are not always very straight forward and require a good amount of understanding in the subject. This becomes quite difficult if you are a working professional to manage time and shift your focus in between contexts. I participated in Analytics Vidhya's Game of Deep Learning Hackathon.

Are you working with image data? This got me thinking -- what can we do if there are multiple object categories in an image? Making an image classification model was a good start, but I wanted to expand my horizons to take on a more challenging task -- building a multi-label image classification model! I didn't want to use toy datasets to build my model -- that is too generic. And then it struck me -- movie/TV series posters contain a variety of people. Could I build my own multi-label image classification model to predict the different genres just by looking at the poster?

It promises to solve your most complicated problems for the small price of an enormous amount of data. The only problem is you are not working at Google nor Facebook and data are scarce. So what are you to do? Can you still leverage the power of deep learning or are you out of luck? Let's take a look at how you might be able to leverage deep learning even with limited data and why I think this might be one of the most exciting areas of future research. Before we discuss methods for leveraging deep learning for your limited data, please step back from the neural networks and build a simple baseline.