Silver Blog, Sep 2017Top 10 Videos on Machine Learning in Finance

Talks, tutorials and playlists – you could not get a more gentle introduction to Machine Learning (ML) in Finance. Got a quick 4 minutes or ready to study for hours on end? These videos cover all skill levels and time constraints!

This ‘Top 10’ list has been created on the basis of best content, and not exactly the number of views. I have also taken special care to walk you through the world of ML in Finance in a gentle, step-by-step manner. To get you motivated, we first begin with talks on the various applications of ML in Finance. Then, to enable access to free financial data, is a video detailing various sources for the latter. To get your hands dirty, we then move on to R and Python tutorials for specific financial use cases. Want to proceed to a comprehensive study of applications of ML in Finance? I have saved the best for last: two playlists that cover the breadth of ML in Finance :) Let’s begin!

1. Talk: How Financial Services Companies are using data and analytics (1.7 K views) - 4 minute talk

In addition to banks using data to make credit and pricing decisions, asset management firms, hedge funds and insurance companies are also embracing the power of data-driven decision-making. Data has become a strategic asset and it is no surprise that roles such as Chief Data Officer and Chief Analytics Officer have emerged in financial institutions.

2. Talk: ML for Credit Risk Assessment (1.3K views) - 38 minute talk

The company Zopa has a Python-based development and modelling pipeline for credit risk assessment. It uses ML algorithms and data visualization to understand the customers. Discussed in this video are its techniques for target definition, feature engineering for the existing 3000 variables, model building and hyper-parameter tuning as well as Zopa’s experience with model stacking and Deep Learning.

3. Talk: Free sources of financial data (2 K views) - 12 minute video

Understand the procedure to navigate the following sites, set filters, and extract historical data from Yahoo Finance and Google Finance, The Federal Reserve Bank of St.Louis, the CBOE, Quandl and the U.S Department of the Treasury.

4. R tutorial: Quantmod R package (10.7K views) - 11 minute tutorial

Use R’s Quantmod package and getSymbols() to pull stocks data. This tutorial also teaches you to display the data on charts and change the charting parameters.

5. R tutorial: Portfolio construction using R (21.3 K views) - 8 minute tutorial

This tutorial teaches you how to download the stock price data for 5 stocks, calculate the efficient frontier and plot stocks in terms of where they sit on the risk-return space. Code here

6. R tutorial: Find the best performing mutual/investment funds in R (3 K views) - 14 minute tutorial

This is a very good tutorial on how to use basic stats in R to find out which are the funds that have performed well consistently over a short, medium and long term. Learn to use a scatterplot to plot the returns over 1 year and 3 years, interpret the graphs and label the companies.

7. Python tutorial: Predicting stock prices in Python (218 K views) - 7 minute tutorial

Siraj Raval’s YouTube videos on ML are a rage. Here, he teaches you 40 lines of code to build a stock-price prediction graph using Python’s scikit-learn. Build 3 predictive models that predict the prices of Apple stock, then plot them on graphs to compare which predictive model performs best.

8. Python tutorial: Predicting stock prices using Deep Learning (161K views) - 9 minute video

It is amazing how Siraj Raval packs a wealth of information in a highly engaging and concise manner. In another video by him, the training data was a time series of daily closing prices of S&P 500 from January 2000- August 2016. He predicts closing prices for a specific day using an LSTM network using Keras with a TensorFlow backend.

9. Playlist 1: (1.9 K views) 42 videos on Machine Learning in Finance (tutorials + concepts + lectures on concepts)

These 40+ hours of video is your go-to resource for the basic concepts in financial trading, interviews with traders, documentaries on stock markets, discussions on backtesting, to ML tutorials to implement various quantitative strategies. The playlist could improve in the area of organization, though, as the videos are in random order.

10. Playlist 2: (72 K views) Python for Finance tutorial series: 24 videos

Saving the very best for last, the ‘Python Programming for Finance’ playlist from sentdex is immaculately well-organized and implements algorithmic trading in Python. A full 10/10 for this playlist. It only involves programming tutorials, though and not videos on understanding finance concepts. However, the playlist above (‘Playlist1’) should help with that :)