Data Encoding Tecniques

Data Encoding Tecniques

1. One hot Encoding

   • Nominal Data
   • For N categories in a variable, it uses N binary variables.
   • Uses 0 and 1 while representing the data.

2. Dummy Encoding

   • Uses N-1 features to represent N labels/categories.
   • Uses 0 and 1 while representing the data.

3. Effect Encoding

   • It is used -1,0 and 1 while representing the data.

4. Binary Encoding

   • Works well when there are a high number of categories.
   • Is a memory-efficient encoding scheme as it uses fewer features than one-hot encoding.

5. BaseN Encoding

   • By default the Base is 2 which is equivalent to Binary Encoding. You can use a larger base such as 4,8 or 10.
   • It uses numerical basis.

6. Hash Encoding

   • Transforms the data in lesser dimensions
   • By default, it uses the "md5" hashing algorithm. Changing the algorithm is up to you.

7. Target Encoding

   • Is a Baysian encoding technique.
   • We calculate the mean of the target variable for each category and replace the category variable with the mean value.
   • Be careful because It can lead to target leakage or overfitting.

8. Ordinal Encoding

   • It imposes an ordinal relationship where no such relationship may exist. This can cause problems and a one hot encoding may be used instead.
   • Ordinal Data