Part 1. Assessing Models with Alternative DataRead the paper: An Intelligent Approach for Predicting Stock Market Movements inEmerging Markets Using Optimized Technical Indicators and Neural Networks bySagaceta Mejia et al.:BOorXs_wOzum9iij5b6EpD-oo_3Dd66uktp0de9a6WA5k8002IveFBelow are five questions from the paper designed to test students’ understanding of thedata, methodology, financial problem, application, and evaluation in the paper. Answerthe 5 questions as detailed in the lists below:Q1. Data Understanding: What types of data are used in the paper to predict stock market movements,and how are technical indicators derived from this data? Discuss the importance of using such indicators in forecasting stock pricetrends.Q2. Security Understanding Pick one of the 3 funds (ECH, EQZ, or IVV). Write a 1-page (strict limit!)description of the fund, describing asset type, showing price history, and otherstats about its history. Why do the authors decide to run a classification problem rather than aregression problem? Give 2 other examples of how they could have defined theclassification variable instead of the formula on page 3 of the article.Q3. Methodology Understanding Separate the 2nd section (2 Materials and Methods) by writing a new section 2called Data. What are the subcategories of this section? (For example, DataProcessing should be one. What are the others?) Call Section 3 Methodology. What are the subcategories of this section? Hint:One should be LASSO. What are the others? How would you divide descriptive statistics from models? (Hint, think aboutPearson correlation versus LASSO). Outline the new section 3 with subcategories. Explain the optimization processof technical indicators used in the paper. How do the authors improve thepredictive power of these indicators, and why is it important to optimize them forthe neural network model?
Leave a Reply
You must be logged in to post a comment.