Internship Blog

Due to pandemic, it was mandatory to stay at home during the lock-down in India. In the beginning of the lock-down-vacation, I scheduled up many productive activities for a month. But, unfortunately, I couldn't follow it for even 15 days. During the time, when I was following the schedule, I came across many websites hosted as portfolio by the people. So, this motivated me, for creating my own Protfolio, and you are even reading this blog on it.

After a few weeks, I received a mail, that CHARUSAT students are offered financial help for like thousands of courses. And also gave us option to do any course, we want, as an internship. So, I wanted learn machine learning from Andrew Ng, as my internship. But unfortunately, that course was not provided among those thousands of courses. So, I applied for financial aid for that Machine Learning course on coursera and waited for 15 days for the approval. After 15 days, thankfully, coursera approved my application and provided financial aid.

Machine Learning, by Stanford University, is a course of 11 weeks. The course also draw from numerous case studies and applications. This course is intructed by ANDREW NG. This course provides a broad introduction to machine learning, data-mining, and statistical pattern recognition.

Topics include:
i) Supervised learning (parametric/non-parametric algorithms, support vector machines, kernels, neural networks).
ii) Unsupervised learning (clustering, dimensionality reduction, recommender systems, deep learning).
iii) Best practices in machine learning (bias/variance theory; innovation process in machine learning and AI).
So, I learnt how to apply learning algorithms to building smart robots (perception, control), text understanding (web search, anti-spam), computer vision, medical informatics, audio, database mining, and other areas.

Every week there are couple of graded quizzes and graded assignments. All the assignments are of applying the algorithm in MATLAB; learned in that week. I even uploaded my assignments on GitHub.

Along with that, Eager Beavers Club's(EBC) website was also under construction. As I was offered a task from the institute to complete that whole website; with a certification portal. Now, this certification portal is a portal which store and validate the certificates of the students. These certificates are those, which students achieve from the events conducted only by the EBC. After the portal gets ready it also needs to be pen-tested before uploading it to the University's main website.

The tasks were:
1) Static, introductory website for the Eager Beavers Club.
2) A certificate portal which:
  i) Stores: The certificates of an individual can only be accessed by his/her CHARUSAT email-id.
 ii) Validation: All the certificates of EBC has a QR-code on it, scanning which you can validate the originality of that certificate.
3) Penetration testing after completion of whole website.

While I was completing the course, I was also working on a project of Handwritten-digit recognition. We know, Deep Learning and neural networks have found use cases in many real-world applications like image recognition, automatic text generation, driverless cars, and much more. So, before diving into these complex areas of Deep Learning, I began with a simple dataset like the MNIST dataset. The MNIST digit classification project is designed to train machines to recognize handwritten digits. For me, as a beginner, its challenging to work with image data over flat relational data. So, the MNIST dataset is best for me. In this project, I used the MNIST datasets to train my model using Convolutional Neural Networks (CNNs). Although the MNIST dataset can seamlessly fit for my PC memory (it is very small), the task of handwritten digit recognition was pretty challenging.

So, summarising the blog, I have 3 projects:
1) My Portfolio
2) Handwritten-digit recognitiion
3) Eager Beavers Club's website

And one training/course:
1) Machine Learning