Regression Analysis and Time Series Analysis

Overview

Conducting predictive analysis on data using Python and statistical methods such as regression and time series.

• Car Price Prediction Dataset
• Netflix's Stock Price

Regression

Building regression models that can be for real world usages.

Steps:

• Fitting OLS Models
• Visualizing Relationships and Linear Models
• Hypothesis Testing
• Plotting Residuals
• More Tests

Time Series Analysis

Conduct time series analysis on Netflix stock price data. The notebook time_series.ipynb contains the research and analysis conducted on the Netflix stock price.

---

Autoregressive (AR) Model

The AR model did not perform well on Netflix's closing price.

Key Notes:

• Transformed the data using Box Cox.
• Subset the data to the last 90-180-365 days.
• Determined performance of the AR model using metrics such as MAPE and MSE.

---

Autoregressive Integrated Moving Average (ARIMA) Model

The ARIMA model was able to perfrom well on Netflix's closing price, but the model might not be able to predict future prices.

Key Notes:

• Transformed the data using differencing.
• Conduct Augmented Dickey-Fuller Test.
• Determine P and Q values using PACF and ACF plots.