Description

The due date is May 3 2020. To receive credit, answers must be prepared in the form of a report of at most 5 pages, 10-12 point font, and e-mailed to me at rpaluszynski@uh.edu. Reports that do not satisfy these conditions will not be considered.

Please attach the data file that you used to make your analysis, but treat it as reference. Do not assume I will be going over it, the report should be self-contained.

Think of this assignment as practice for creating professional reports in your future job. Put some effort into writing and editing it nicely. In particular, grammar and spelling errors will be penalized. Make sure your graphs are legible and clearly labeled.

Download the dataset of financial crises classification that accompanies our course textbook “This Time is Different” from reinhartandrogoff.com.1 Choose a group of countries that interest you from the list. For example, you could select a geographical group (Western countries, Latin America, Asia, etc.), or choose them by development level (poor, rich, middle-income, etc.). You can also use all of them if you want. Think about which type of financial crisis you would like to predict for your group (currency or inflation crises, stock market crashes, domestic or external sovereign default crises, banking crises).

Next, go to the website of World Development Indicators (data.worldbank.org/products/wdi) and click on “Query and analyze data in databank”. Check the same countries as you chose before, with the time frame 1960-2010. Pick 15-20 data series (or more if you like) that you think are relevant in predicting financial crises. Make sure to choose a mix of level and change variables (for example GDP per capita vs. GDP growth rate). Download the data into an Excel spreadsheet.

i. Report the countries, the data series, and the crisis type that you chose. 

ii. Combine the two datasets and transform them into a single one, where each observation represents a country-year. Make sure that observations are in rows and variables are in columns. It is useful to have the first row contain variable names. Save this spreadsheet to a CSV file. 

iii. Load the dataset into Matlab using the command readtable, for example:2 tab=readtable(’wdi.csv’); 

iv. Once you have loaded the data into Matlab, go to Apps → Classification Learner. Open a new session, choose your workspace (the data table you loaded), the response