Faculty Development Program On the Deep Learning : May 27-May 31, 2019
Day - 1
Dr. Sumit Srivastava, Professor, IT, Manipal University has started the FDP with the discussion on Deep learning and its application in a different area of the engineering. He explained the basic architecture of an artificial neural network to the latest architecture of Deep learning model nowadays. Further, the discussion covered the future of deep learning and the ideas where the participants can approach for the research projects. He covered the mathematical foundation of the artificial neural network. He has also explained the importance of and use of optimizers to tune the model perfectly. During the session, the participants also enthusiastically participated to ask the query and also clear their doubts on Deep learning. The participants practically got to know the working of artificial neural network and deep learning in hands-on session.
Day - 2
Dr. Mahesh Jangid, Assistant Professor, CSE, Manipal University has discussed two deep learning approaches called CNN and RNN. He firstly explained, “Why deep learning is so popular among the research community nowadays”. Then he took the (CNN) Deep convolutional neural network and explained the working of convolution, pooling, dropout and fully connected layers. He has done many hands-on session on MNIST dataset, object detection, CIFAR dataset, etc. Further, he explained the role of bias and variation to detect the underfitting and overfitting problem of deep learning models. To work on time series dataset, he explained the RNN (Recurrent Neural Network). The gradient vanishing and exploding problems were also discussed during the session. After the lunch session, he took a hands-on session on the real-time stock prediction and the image captioning using the CNN and RNN models.
Day - 3
Dr. Sandeep Chaurasia, Associate Professor, CSE, Manipal University gave the lectures on natural language processing (NLP) on the third of FDP. He firstly introduced the foundation concept of Natural Language Processing along with the state of art research areas in NLP. Then he explained the NLTK toolkit & Machine learning using text data. Further he explained the sentiment analysis for movie dataset and exploratory data analysis was done for analysing the dataset. He discussed the Markov chain model for generating the automated test and Liner Discrimination Analysis (LDA) for identifying the topic. He gave many demonstration to the participants to get in-depth knowledge of the natural language processing (NLP).
