Machine Learning vs Deep Learning

When it comes to AI and Data Science domain, people are often willing to know what’s the difference between Classical Machine Learning and Deep Learning. And I would say that its quite important because both are 2 different areas of AI with completely different approach in solving the same/similar problems.

On one side, Classical Machine Learning is quite easy to start with if you know R or Python with basic knowledge of the actual Algorithm or Mathematics behind it. Its based heavily on the Statistics as its uses:

• descriptive stats
  • Measures of Central Tendency
  • Measures of Dispersion
• inferential stats
  • Parametric Tests like T-Test, Z-Test, Pearson Correlation
  • Non-parametric Tests like Chi-Square, Spearman and Kendall Correlation

Whereas, Deep Learning needs deeper knowledge on the Neural Networks with the understanding of the types of layers [dense layer, convolution layer, etc. ] and the Activation Function [sigmoid, ReLU, etc.] that you select for the problem solving.

Knowledge of DL frameworks like Keras, Tensorflow, torch, caffe(2), theano, CNTK play a vital role in the successful implementation of a DL model.

However, with the increased Complexity of Deep Learning, you loose the Explainability for the end user.

To start your Data Science Journey, start with this Page
 today...