Find Undervalued Stocks Using AI & Python

---

---

AI Summary

Video Summary: Predicting Stock Value with Machine Learning

• Introduction
  • Welcome to a video on Python, AI, and machine learning.
  • The goal is to predict if a stock is undervalued using machine learning.
• Setting Up
  • Use Google Colab (collab.research.google.com) for Python programming.
  • Log in with a Google account, create a new notebook, and start coding.
  • Encouragement to subscribe and like the video for channel support.
• Importing Libraries
  • Import necessary libraries in a new cell:
    • pandas as pd
    • RandomForestClassifier from sklearn.ensemble
    • accuracy_score from sklearn.metrics
    • train_test_split from sklearn.model_selection
  • Run the cell to check for errors.
• Loading Data
  • Upload a CSV file containing stock data.
  • Rename the file to data.csv.
  • Load the data into a variable and display its contents.
  • The data includes columns for ticker symbol, PE ratio, current price, earnings per share, fair market value, over/under value, and value percentage.
• Preparing Data
  • Split data into features (X) and target (Y) datasets.
  • Features include PE ratio, current price, and earnings per share.
  • Target is the column indicating if a stock is undervalued.
• Splitting Data
  • Split the data into 80% training and 20% testing datasets using train_test_split.
• Creating and Training the Model
  • Create a RandomForestClassifier model with 100 estimators and a random state of 0.
  • Train the model on the training datasets.
• Making Predictions
  • Predict values on both training and testing datasets.
  • Store predictions in separate variables for comparison.
• Evaluating Accuracy
  • Calculate and print the accuracy of the model on both training and testing datasets.
  • Training accuracy is 100%, testing accuracy is around 95%.
• Comparing Predictions
  • Show the model’s predictions and actual values from the test dataset to compare.
• Using the Model
  • Demonstrate how to use the model to predict the value of other stocks given their features.
  • Use an example stock and its features to predict if it’s fair or overvalued.
• Conclusion
  • Reminder that the model is simple and not perfect.
  • Encouragement to do further research before investing.
  • Thanks to viewers and Patreon supporters.
• Support and Code Access
  • Patreon support and access to code and dataset mentioned.
  • Patreon link: patreon.com/computerscience provided for support.