Kyso Weekly: Issue #9
Remember to both star and share the studies you really like!
This post is part of an ongoing Kaggle competition, the aim of which is to classify audio files that cover real-world sounds from musical instruments, humans, animals, machines, etc. Trevor takes the reader through EDA, then constructs a 1DCon Keras model on raw audio waves, followed by 2DCon on MFCC (Mel Frquency Cepstral Coefficients), one of the techniques used before the Deep Learning era to extract features from audio signals. Which model performs better??
In mathematics, Buffon's needle is the earliest problem in geometric probability to be solved, a question first posed in the 18th century by Georges-Louis Leclerc, Comte de Buffon. The problem is to randomly drop a needle on the unit square and count the number of times that the needle touches (or "cuts") the edge of the square. This notebook illustrates some basic concepts of Monte Carlo methods, showing how to code up a quick simulation that visually reveals the key issues and then how to boil this down into a simulation to compute the estimated solution.
Portfolio optimization is a mathematically intensive process that can be accomplished with a variety of optimization functions that are freely available in Python. This post is an exellent guide for those working with financial data, as the author takes the reader through each of the following steps:
- Building a function to fetch asset data from Quandl.
- Converting the data into return and covariance matrices.
- Simulating a basket of thousands of portfolios using the same assets.
- Showing how portfolio weights can be optimized for either volatility, returns, or Sharp Ratio.
- Building the Markowitz efficient frontier.
- Building the Capital market line.
- Calculating the optimal portfolio weights based on the intersection of the capital market line with the efficient frontier.
That is all for now!! Have a great week!!