Speaker: Bhanu Prasad Kotamraju
Advisor: Prof. Dominique Durand
Title: Unlocking the Power of Machine Learning: A Beginner's Guide to Time Series Classification and Feature Extraction in Neural Data
Abstract: The objective of this presentation is to empower individuals in effectively applying multi-layer perceptron (MLP) models to their research problems by providing comprehensive guidance on approaching specific research problems and datasets. By the end of the session, attendees will possess the necessary knowledge to readily implement MLP models and adapt them to their research endeavors.
The presentation explores various terminologies associated with machine learning (ML) and provides a fundamental understanding of basic ML algorithms, such as the perceptron. It delves into the implementation of an MLP model, specifically focused on its application for time series classification using neural data. Furthermore, the critical aspects pertaining to feature extraction for the algorithm are discussed, offering valuable insights into interpreting these features. In addition, a brief overview of more advanced models, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and long short-term memory (LSTM) networks, will be provided.
Moreover, the presentation delves into recent trends in deep learning, shedding light on how researchers are leveraging these advances to effectively tackle challenges in the field of neural engineering. This discussion provides attendees with valuable insights into the evolving landscape of deep learning and equips them with knowledge on incorporating these advancements into their research projects.