Predict Stock Prices

Stock Market Prediction using Machine Learning

“With a good perspective on history, we can have a better understanding of the past and present, and thus a clear vision of the Future.”

— Carlos Slim Helu

Photo by Aditya Vyas on Unsplash

Introduction

Stock market analysis and predictions are very difficult tasks to complete. There are numerous variables that influence the value of stock prices including, but not limited to, media publications, insider buying, notice of dividends, stock splits, quarterly financial reports and financial forecasts by management. These variables make it extremely difficult for any stock market professional to predict the rise and fall of stock prices with accuracy.

With the introduction of Machine Learning and its many advanced algorithms, the most recent market research and stock market prediction advancements have begun to include machine learning approaches in analyzing stock market data and predicting future stock prices. The successful prediction of a stock’s future price could yield a significant return on investment.

Disclaimer: this article is not intended to provide trading and investment advice. No warranties are made regarding the accuracy of the model. Readers are advised to due their own research before making any investment decisions.

Roadmap

The remainder of this article will present a program that will attempt to predict future stock prices based on previous closing prices using machine learning. We will use the linear regression model. The following steps will be performed using Python.

Install the packages.

Import the libraries.

3. Import the dataset into a dataframe.

4. Analyze the data.

5. Restructure the data.

6. Split the data into training and test data sets.

7. Train the model on the training data.

8. Make predictions on the test data.

9. Visualize predictions for stock prices.

The Program

The objective of this program is to predict future stock prices based on previous closing prices for a company stock. In this case we will use Tesla.

We will start by installing the required packages and importing the libraries.

Install the Packages

# Install yfinance and cartopy packages. !pip install yfinance !pip install cartopy

Import the Libraries

# Import the libraries. import cartopy import pandas as pd import yfinance as yf import datetime import numpy as np import matplotlib.pyplot as plt

Import the Data Set

We will use Yahoo Finance Market Downloader to get stock data for Tesla for the time period from 1/1/2021 thru 12/5/2021.

# Get stock data for Tesla. tesla = yf.download('TSLA',start='2021-01-01', end='2021-12-05',progress=False)[['Close']]

Data Analysis

Let’s take a look at the data and show some information about the data elements.

# Print the first 5 rows. tesla.head(5)

# Show information about the data. tesla.info()

<class 'pandas.core.frame.DataFrame'> DatetimeIndex: 233 entries, 2021-01-04 to 2021-12-03 Data columns (total 1 columns): # Column Non-Null Count Dtype --- ------ -------------- ----- 0 Close 233 non-null float64 dtypes: float64(1) memory usage: 3.6 KB

There are 233 stock records for Tesla within the specified time frame. There is a Date column (index) and a column named Close which has a float data type. The data set contains no null values.

Restructure the Data

We need to transform the data so we can feed the data into the machine learning model. We will restructure our data so that each row contains the actual stock price for one day followed be each of the closing stock prices for the previous 10 days. For example, beginning with the 11th row, we will get the previous 10 closing prices and use them for the features. The Pandas shift function will be used to get the close values from each of the previous 10 rows of data. This process will be repeated for each row of data.

# Restructure the data. tesla = pd.concat([tesla, tesla.shift(), tesla.shift(2),tesla.shift(3),tesla.shift(4),tesla.shift(5), tesla.shift(6),tesla.shift(7),tesla.shift(8),tesla.shift(9),tesla.shift(10)], axis=1).dropna() tesla.columns = list(range(0,11)) tesla.rename(columns={0:'actual_stock_price'},inplace=True) tesla.head()

Listed below is an image of the dataset after the data has been restructured.

Split Data into Training and Test Data Sets

# Split the data into training and test data sets. train = tesla.head(len(tesla)-10) test = tesla.tail(10)

Train the model and make predictions on the test data

We will use the Linear Regression model for training and to make predictions on the test data. Linear Regression is a supervised machine learning model in which the model finds the best fit linear line between the independent and dependent variables.

# Import the Linear Regression model from scikit-learn. from sklearn.linear_model import LinearRegression # Create the model object. LR = LinearRegression() # Train the model on the training data. LR.fit(train[list(range(1,11))],train['actual_stock_price']) # Make predictions on the test data. test['predictions'] = LR.predict(test[list(range(1,11))])

Visualize the Actual and Predicted Stock Prices

Let’s visualize and compare the actual stock prices and the predicted stock prices over the specified time period.

# Visualize the actual stock price and the predicted stock price over the time period. plt.figure(figsize=(12,7)) ax = test[['actual_stock_price','predictions']].plot(figsize=(12,8)) plt.xlabel('Date', fontsize=12) plt.ylabel('Stock Price', fontsize=12) plt.title('Actual and Predicted Stock Prices', fontsize=15) plt.grid() plt.show()

Conclusion: In general, the predicted stock prices and the actual stock prices show similar trends in direction. Many times the predicted stock prices are below the actual stock prices. On 11-23-2021 the actual stock price went down and the predicted stock price went up. On 12/03/2021 there was a significant drop in the actual stock price and a moderate decrease in the predicted stock price.

Thanks so much for reading my article! If you have any comments or feedback please let me know.

If you enjoy reading stories like these and want to support me as a writer, consider signing up to become a Medium member. Membership gives you unlimited access to stories on Medium. You can sign up using this link https://medium.com/@dniggl/membership