class: center, middle, inverse, title-slide .title[ # Group 5: Adverse Event Classification in 8-K Filings ] .author[ ### Abhishek Varma, Kevin Roy, Saurabh Shete ] --- ## What is 8-K filing and EDGAR? * 8-K filings are reports of unscheduled material events or corporate changes at publicly traded companies. * 8-K filings, submitted to the U.S. Securities and Exchange Commission (SEC) through its Electronic Data Gathering, Analysis, and Retrieval (EDGAR) system, are specifically designed to report on major events that shareholders should be aware of. * They provide valuable insights into the financial health and stability of a company. * Investors, regulators, and the companies themselves rely on 8-K filings for informed decision-making. --- # Problem Statement ## The Challenge of Adverse Event Classification * Adverse events (e.g., product recalls, regulatory actions, financial restatements) can significantly impact a companys performance and reputation. * Manual classification of adverse events from 8-K filings is time-consuming and subjective. * An automated, objective approach is needed to classify the severity of adverse events. --- # Proposed Solution ## Leveraging Deep Learning for Automated Adverse Event Classification * We propose using a deep learning model to classify adverse events from 8-K filings. * We will extract relevant features from the 8-K filings and metadata using NLP techniques. * Our model will be trained on a dataset of 8-K filings obtained through the EDGAR API. * We will use H2O/Keras libraries for deep learning. --- # Project Plan ## Key Steps and Evaluation Metrics * Data preparation and cleaning * Exploratory data analysis (EDA) * Feature engineering * Model selection and training * Evaluation using precision, recall, F1-score, and AUC