Tuesday, May 30, 2017

Machine learning is the future

I am very enthusiastic about machine learning and the potential it has solve tough problems. Considering the fact that the amount of data we produce digitally is growing geometrically as online activity och number of people connected to the internet keeps on increasing. We have to make something good out of that massive information we generate.

That is where machine learning come in. Machines are faster and can handle and processing large amounts of data better than us humans. Here are some of the things machiné learning can achieve today. It can identify images, write music and even stories give an image, It can write rap lyrics. It can predict the stock market better that any human, generate choreography based on the music and so on and forth. We are just scratching the surface in AI.

Due to my interest in this subject, i started watching some videos about machine learning and focused my attention on deep learning which uses neural networks . Most of the implemention  is in the Python language so i started learning python because most of the libraries and tutorials i see online are all written in python and is it an expressive language. If i want to be part of or become a data scientist and learn faster, I need to start writing some python code which i personally think it will be fun.
 
One technique i machine learning is using Neural networks based on how we think our brain takes in and processes information . Neural networks are incredibly powerful The good news is some of the maths subjects I learned in University like calculus, linear algebra and statistics all come together to produce this powerful technique of doing machine learning .I had to brush my skills with the, matrices, derivatives, gradient descent, sigmoid functions ,probabilities and back propagation . It sets a good base to understand what going underneath these python libraries . One can just focus on the building the models and choosing the techniques to handle the data.

Becoming a data scientist is more of brushing some of your maths skills in statistics, linear algebra and calculus . You must be capable to clean, understand and model data and to write code that turns that data to something useful. There are a lot of resources out there. I will start with small datasets about my life and then later handle complex datasets as I get better in this way of problem solving. I think I am in for a treat. It will be fun and a challenge to use and play with data to so some good for humanity.

It is time to train some datasets and solve some real tough problems

Happy machine learning !
Ngala Talla

No comments:

Post a Comment

Machine learning is the future

I am very enthusiastic about machine learning and the potential it has solve tough problems. Considering the fact that the amount of data we...