beetrandahiya Goto Github PK

followers: 17.0 following: 1.0 repos: 59.0 gists: 0.0

Name: Prakrisht Dahiya

Type: User

Bio: I'm Prakrisht, a visionary web developer and python programmer studying Quantum Computing .

Twitter: PrakrishtD

Location: Leipzig, Germany

Blog: https://prakrisht.me/

PhiloMath
Visionary Web Developer
Python Programmer studying Quantum Computing.
and A bit of an Artist as well

Connect with me:

Languages :

Frameworks :

Libraries :

Prakrisht Dahiya's Projects

3d_plot_html

yea, trying some 3d plotting thingy with plain js

beet_chat

This is a chat app, based on React and Firebase.

bmi_calculator

This is a basic webpage, where you can find your BMI (Body Mass Index) and you might also find some nice suggestions.

This is my first nice machine learning model, This model gave a 97.85% accuracy in classifying between Cats and Dogs. I made it using a pre-trained base model MobileNet V2 , and after that i added a global average pooling and then a dense layer for categorization between two classes ( cats and dogs) , i used only one dense neuron in last layer even when i had to classify 2 classes because , as we know that in test data there will only be two type of images cats and dogs , so we can just classify dogs and then ones that doesn't qualify the classification are cats. It has been trained on a small dataset to show how a machine learning model can work well even on a small dataset if we use pre-trained models inn architecture.