Ahmed-M-G/House-Prices-Advanced-Regression-Techniques-DA

Using an advanced regression techniques to predicte the house price

House-Prices-Advanced-Regression-Techniques-DA

EDA, Feature Engineering, and AutoML

• The goal of this assignment is to analyze 3 years’(2018-2020) house sales data provided by New York City (NYC) goverment and build regression model to predict house price. NYC has five boroughs, i.e., Bronx, Brooklyn, Manhattan, Queens and Staten Island. Sales of houses in each borough has been provided.

Task 1: Read dataset, merge data, and perform basic data exploration. Note, you should briefely discuss the quality of dataset (missing values, duplicate entries, etc.) Also, in task 3, you would be asked to perform prediction on house sale price, thus it would be good to consider outliers related to this prediciton task when you perform data cleaning.

Task 2: Data exploration using data visualization. Raise two questions that can be answered by performing data visualization. Briefely mention why you think this question would be interesting to whom (who is your audience). Think about the EDA principals.

Task 3: Data Exploration via Statistical Test Raise one question that can be answered by performing hypothesis test. Briefely mention why you think this question would be interesting to whom (who is your audience). Also mention which statistical test you would choose and why.

Task 4: Feature Engineering. If we would like to predict the house sale price. Analyze the scale of each attribute and determine which ones you would transfer (e.g., cateogorical features). Discuss how you plan to select important features.

Task 5: AutoMLUsing Auto-sklearn to explore performance of one state-of-the-art autoML tool on the given data (after your previous preprocessing). Optional: compare with performing autosklearn on raw data.