FAANG Stock Forecasting: Using Ridge Regression, Linear Regression, and Neural Networks.

Abstract

Blue-chip stocks serve as an important cornerstone in the stock market especially for long-term investors, since they specialize in high stability, high liquidity, and long-term growth. However, predicting future prices remains a challenge because of increased market volatility. This research aims to compare and evaluate the efficacy of three machine learning models: a ridge regression model, a linear regression model, and a neural network model— in order to forecast top-performing tech stock prices and trends over the long term. Using historical data (opening/closing prices, high/low prices) from Meta (formerly known as Facebook), Amazon, Apple, Netflix, and Alphabet (formerly known as Google), each of the three types of ML models will be developed for each stock. The models will all be separately trained and tested; they will be assessed for predictive accuracy using various success metrics (MSE, R2, RMSE) and will be compared with each other using MAE as the common success metric. The neural networks had the MAE with the least value, with ridge regression having the greatest MAE value. Based on this MAE comparison, the research concluded that the LSTM recurrent neural networks had the most accurate outputs with minimal error, linear regression performed the second best and ridge regression performed the worst.