So this is my second blog, after almost two and half months since I joined and published my first article on the platform. As you know, I am a data enthusiast learning machine learning and deep learning, I started learning machine learning and AI almost three or four months ago. I learned many things in this period of time. Today, I thought about writing an important lesson that I learned during my learning phase of machine learning and AI.

Although I am a technical student coming from engineering background, I come from a domain other than computer science (i.e., mechanical engineering). I learned very few programming concepts with C at college. Of course this helped to understand some basic concepts to start off with Python for data science, machine learning and AI. Maths is very much needed to understand various machine learning and deep learning algorithms, that are the main heroes behind the success of AI today. I knew some maths from high school, that would work just fine for learning how machine learning and deep learning algorithms work.

The internet is full of information. It is often very difficult to find the content that would help the most to learn any skill and be good at it. This was the first problem I encountered when trying to learn machine learning and AI through self study. I would often get lost in the vast source of contents available in internet. While scrolling through the twitter feed, I found a roadmap to learn machine learning and deep learning posted by a @svpino. He shares various machine learning and AI contents on twitter. So I started to follow that roadmap. Deep Learning Specialization by DeepLearning.AI on coursera is one of the part of that very roadmap I was following.

So, this specialization is all about deep learning and various neural network architecture due to which the AI today has made tremendous progress in various fields. The last two courses on this specialization were somewhat harder for me to understand. I could not understand how I was going to implement all these in code. I somehow managed to understand CNN and RNN architecture and completed the specialization through some help from YouTube tutorials. After completing it, I started learning Tensorflow Developer Professional Certificate in coursera. During this professional certificate, I learned how to implement all those CNN and RNN architectures using high level APIs of tensorflow and keras. All those stress I have had before with implementation of deep learning algorithms, were gone. I became full circle with theory from previous specialization course and implementation strategy on this professional certificate.

I think it's important for us to think and see only the very next step, for the moment rather than thinking about the whole path and our target. As long as we can see what and where our next step should be, it is a waste of time and energy to get stressed about not being able to see the complete path right away. With all those next steps, we eventually get close to our target and at the end we can learn and see the whole path completely even if it is unclear for now. Such paths are never straightforward and are often difficult to view. So just take that next step!