Machine Learning Made Easy
May 25, 2018
Machine Learning Made Easy
Score 8 out of 10
Vetted Review
Verified User
Overall Satisfaction with MATLAB
I initially purchased MATLAB 2017b for a mathematics course during graduate school. At first, I was skeptical about the product and why it was being used. However, I quickly discovered the features that make MATLAB stand out from C++ and Mathematica. MATLAB truly makes numerical analysis easier. The ability to easily push code to run on a GPU allows for a user to take advantage of the modern day hardware. The signal analyzer, filter builder, and machine learning apps have saved me and my co-authors hours by allowing us to quickly look at the effectiveness of a filter on raw data. The built-in K-NN and Neural Network algorithms were easy to use and produced fantastic results for our specific project.
- Quick visualization of data allows the user to improve code as its written.
- Easy to use syntax makes picking up MATLAB quick.
- Constant improvement and addition of apps and functions keep MATLAB competitive with other languages.
- Errors are simple to understand and fix.
- Improvement to the symbolic solution output for mathematics would be huge as the output is currently ASCII art.
- Made collaboration with colleagues quick and painless.
- The publication features reduced need to produce formal LaTeX write-ups.
- Forced some computers to show their age.
For working with experimental data MATLAB takes the cake. The ability to quickly import and plot raw and processed data was the selling point for MATLAB. However, MATLAB is not well suited for intense symbolic evaluation of mathematical problems. My gold standard for symbolic solutions is Wolfram Mathematica.