# Machine Learning Algorithms

Because of so much hype about deep learning in recent years, people are trying to go for DL based approaches. But, you can get better results with basic machine learning algorithms if you have fewer data points. In [this](https://arxiv.org/pdf/1509.01626.pdf) paper, they compared traditional ML algorithms with deep learning algorithms. You can check the results below

![https://arxiv.org/pdf/1509.01626.pdf](https://537411571-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-LszrzsSXnrNcKntCqrn%2F-M5b0RKjyze8oB5hSpA8%2F-M5bHUOAwMwsxJmzl7bO%2Fimage.png?alt=media\&token=c35b178b-432d-4d55-8703-008fd06bf473)

From above image, they got better result on test set for first four `datasets` using traditional ML methods. First four `datasets` contains fewer data points compared to next four, you can check that below. 

![https://arxiv.org/pdf/1509.01626.pdf](https://537411571-files.gitbook.io/~/files/v0/b/gitbook-legacy-files/o/assets%2F-LszrzsSXnrNcKntCqrn%2F-M5b0RKjyze8oB5hSpA8%2F-M5bRtq0U2-eXZSkTvEZ%2Fimage.png?alt=media\&token=fe65b653-a421-4734-beec-ffc4dbd1607d)

You can also `interpret` your models clearly if you are using some ML-based techniques. Based on your problem, business constraints, and the size of the data you have, you can choose the appropriate technique. On the contrary, In recent years, we got **transfer learning** and active learning concepts for text processing, using these, you can achieve better results for fewer data point `datasets`. 

In the next post i will take a algorithm, try to create different types of features to train and, try to summarize all other algorithms. 

References:

1. <https://arxiv.org/pdf/1509.01626.pdf>


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://udai.gitbook.io/practical-ml/natural-language-processing/text-classification/ml-classification-algo.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.