The assignment is structured into three phases, where students are required to analyze and interpret stock price data for three different companies. One of these companies must be the one you currently work with, or$,$ if it is not listed, a listed company that closely resembles your company. The other two companies should be from different industries, categorized as large, medium, and small companies, all listed on any exchange. The objective is to execute the provided code for each company, analyze the results, and write detailed interpretations and comparisons. $0\mathrm{.}5$ Marks will be deducted for every $1$ day of delay and $0\mathrm{.}5$ mark will be deducted for every additional page beyond $2$ pages of Analysis

Phase $2$: Time Series Analysis, ML Model Development, and Model Identification $(10$ Marks$)$

Objective: Explore time series characteristics, apply ML models, and prepare data for forecasting.

Stationarity Tests $($ADF and KPSS$)(1$ Mark$)$

Perform ADF and KPSS tests on each company$$s time series data to assess stationarity.

Discuss the importance of stationarity in time series analysis and interpret the test results.

Differencing and Transformations $(1$ Mark$)$

Apply differencing, log transformation, and detrending techniques to achieve stationarity.

Explain the impact of these transformations on the time series.

Time Series Decomposition $(1$ Mark$)$

Decompose the time series into trend, seasonal, and residual components using additive and multiplicative models.

Compare and discuss the decomposition results across the three companies.

Autocorrelation and Partial Autocorrelation $(1$ Mark$)$

Generate and interpret ACF and PACF plots for each transformed time series to identify significant lags.

Explain the role of these plots in model selection.

ARIMA Model Fitting $(1$ Mark$)$

Fit ARIMA models to each company$$s data using identified parameters.

Compare the model fits and discuss any notable differences in forecast accuracy.

Forecasting with ARIMA $(1$ Mark$)$

Forecast future stock prices using the fitted ARIMA models.

Interpret the forecast results and discuss how they align with historical trends.

Machine Learning Models $($Decision Trees, Random Forest, XGBoost$)(1$ Mark$)$

Apply decision trees, random forest, and XGBoost models to predict stock prices.

Discuss the performance of these models and compare them with traditional ARIMA results.

Model Evaluation and Comparison $(1$ Mark$)$

Evaluate the ML models using performance metrics such as MAE and RMSE.

Compare the predictive performance of ARIMA and ML models and discuss any observed advantages or limitations$.$

Feature Importance in ML Models $(1$ Mark$)$

Analyze feature importance from the ML models to understand which variables are most predictive.

Discuss how feature importance insights can influence investment decisions.

Insights from ML and Time Series Models $(1$ Mark$)$

Summarize key takeaways from the ML and time series models.

Provide a comparative analysis and discuss potential strategic applications of each modeling approach.