I know I did not put out 6 days of the challenge. This challenge is hard and trying to publish every day is even harder, so expect some days where there might not be any updates. But, I am trying to be consistent.

So, let me go on with the updates,

Daily task 1: Deep learning

Today I was able to complete a lecture on Linear classifiers and it was one of the best lectures that I've seen about them. The lecture was about linear classifiers for the first half and then it was about loss functions / objective functions. Some important points that I've learned today are:

• Basically, linear classifiers are like the most basic parametric approach, where we just multiply a matrix and an input vector 'x' ( we also add a bias as an offset)

• These can be interpreted in 3 different ways and with each of these interpretations, some features or problems of them will be quite evident

• we have : algebraic viewpoint, visual viewpoint, geometric viewpoint

• Multiclass SVM loss and Cross entropy loss, are both used for multiclass classification and their mathematical details

• Regularization

I plan to make a separate article explaining all of these things, so I am not writing anything here.

Daily task 2: 180 days of code

Today I learnt arrays in C++ and how they are stored in memory and the basics of those. I also coded how to get the min and max value for a given array

#include <iostream>
#include <climits>
using namespace std;

int main()
{
    int arr[5]= {1,3,4,5,6};

    int ans = INT_MAX;   // most important part

    for(int i=0; i<5; i++)
    {
        if(arr[i] < ans)
        {
            ans = arr[i];
        }
    }
}

Using INT_MAX and INT_MIN to code this is the little trick here !!!

Daily task 3: math

I couldn't do anything from this today, so I will try to cover this tomorrow, hopefully.

This was it for the day and I will try to be consistent and subscribe for updates or if u want to be an accountability partner ( comment too, so that i will know lol )