Data mining
======================================================================
Here we have a lot of information about different topics.


Slides
----------------------------------------------------------------------------------------
.. list-table:: Data mining.
    :widths: 10  20 
    :header-rows: 1

    * - Model
      - Material
    
    * - Clustering 
      - `Slides clustering <https://github.com/it-ces/PUBLIC-AI/blob/main/ML-models/Cluster-slides.pdf>`_
    * - Association rules (Apriori)
      - `Slides Apriori algorithm <https://github.com/it-ces/PUBLIC-AI/blob/main/APRIORI/2.%20A%20priori.pdf>`_
    * - Principal Components Algorithms(PCA)
      - `Slides PCA <https://github.com/it-ces/PUBLIC-AI/blob/main/PCA/PCA_slides.pdf>`_

In this section we will find information about different topics as 
Association rules, clustering and, reduction of dimensionality.


Laboratories
***************************************************************************

* `PCA - linear regression <https://github.com/it-ces/PUBLIC-AI/blob/main/PCA/PCA(ai_page).ipynb>`_
* `Course of dimensionality simulation <https://github.com/it-ces/PUBLIC-AI/blob/main/PCA/Curse_of_dimensionality_simulation(ai_page).ipynb>`_
* `Clustering <https://github.com/it-ces/PUBLIC-AI/blob/main/CLUSTER/clustering-K-H-DS(ai-page).ipynb>`_




Apriori algorithm
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

We are interested in find out if the products $X$ and $Y$ are sistutes  or complementaries.
in a database with transactions made it of inovices in a period of time:

+----+--------------------+
| id |      Products      |
+----+--------------------+
|  1 |  Milk, Bread,...,  |
+----+--------------------+
|  2 |   Eggs, Coce,...,  |
+----+--------------------+
|  3 | Milk, Cookies,..., |
+----+--------------------+
|  . |         ...        |
+----+--------------------+
|  N |  Oranges, lettuce  |
+----+--------------------+


How to known what products will be bought togheter of a total of $N$?
if we assest each possible combination then will have a computational complexity of 

.. math::
    :nowrap:

    \begin{equation}
        3^{N} - 2^{N-1} + 1
    \end{equation}

Therefore the Apriori algorithm allow us solve this in a more effcient way.


Clustering
----------------------------------------------------------------------------------

The unsupervised learning is a field of learning.


K-means
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~



.. warning::

    The models based in distance are affected by `curse of dimensionality`
    to see more information about that please refer to.