Parameters vs hyperparameters in machine learning

In this short video we will discuss the difference between parameters vs hyperparameters in machine learning. Parameters is something that a machine learning model trains and figure out such as weights and bias for the model. Hyper parameter on the other end is something you manually specify such as number of hidden layers, neurons in each hidden layer, epochs etc.

#️⃣ Social Media #️⃣

❗❗ DISCLAIMER: All opinions expressed in this video are of my own and not that of my employers'.

[codebasics]

There was a typo in the presentation. 0:36 it is quite opposite. intercept_ and coef_ are actually parameters and the C, gamma, kernel etc are hyper parameters.

[sandeepr]

Hi, @ 0:36 I think model.coef_and model.intercept_ must be parameters because they are learnt during the training.
Similarly C, gamma and kernel must be Hypermaters as they are user specified. Correct me if am wrong.

[SayyidahHumairah]

Super clear & concise explanation. Thank you!

[SuhaibSalah-ow]

Many thanks for the short, simple and clear video.

[YasirNiazKhan]

In the second example of SVC, you switched the terms parameter and hyperparameter.

[sathyaprakash]

Sir please check @0:36 it is quite opposite

[RahulParmar-ldut]

Dhaval bhai, pls make video on Maths behind Empirical Risk Minimization

[abhi_and_seki]

Perfect!! Exactly what I was looking for :)

[nikt]

It is easy to switch carrier from big data analyst to data scientist ?

[qw]

Hi there I have some experiment data, can I ask u the CNN in the real work ?

[ashishgarg]

Can you pls make playlist on multimodal data fusion for model??

[kirandeepmarala]

Sir Really Enjoying these type of Short Video's... please Continue.. Thank you very much

[khan]

Sir really explaination is very helpful in all the topics ...sir plz create a vedio on library management system using barcode scanning entry project using Django project

[myown]

I'm doing bachelor economics. But unfortunately I'm weak in mathematics.( I don't like maths ) So what's the chances that I can become data analyst. Or you can suggest me. What career is good for me. ?

[rudrakshya]

For Data Science why you use Windows not a Linux Distribution like Ubuntu. I heard that Ubuntu has more support than Windows. Thank you.