new version test

Author: LearningAlgorithms

_config.yml

@@ -7,33 +7,33 @@
 
 # jekyll-seo-tag settings › https://github.com/jekyll/jekyll-seo-tag/blob/master/docs/usage.md
 # --------------------------
-title: Chirpy                          # the main title
+title: Learning Algorithms                          # the main title
 
-tagline: A text-focused Jekyll theme   # it will display as the sub-title
+tagline: A blog about Maths and ML   # it will display as the sub-title
 
 description: >-                        # used by seo meta and the atom feed
-  A minimal, portfolio, sidebar,
-  bootstrap Jekyll theme with responsive web design
-  and focuses on text presentation.
+  A new blog about how mathematics 
+  are the base of Machine Learning
+  and introductory examples of ML.
 
 # fill in the protocol & hostname for your site, e.g., 'https://username.github.io'
-url: ''
+url: 'https://learningalgorithms.github.io'
 
-author: your_full_name                  # change to your full name
+author: Carlos Marco                  # change to your full name
 
 github:
-  username: github_username             # change to your github username
+  username: LearningAlgorithms             # change to your github username
 
 twitter:
-  username: twitter_username            # change to your twitter username
+  username: JC_MarcoG            # change to your twitter username
 
 social:
-  name: your_full_name                  # it will shows as the copyright owner in Footer
-  email: example@doamin.com             # change to your email address
+  name: Juan Carlos Marco                  # it will shows as the copyright owner in Footer
+  email: thelearningalgorithms@gmail.com             # change to your email address
   links:
     # The first element serves as the copyright owner's link
-    - https://twitter.com/username      # change to your twitter homepage
-    - https://github.com/username       # change to your github homepage
+    - https://twitter.com/JC_MarcoG      # change to your twitter homepage
+    - https://github.com/learningalgorithms       # change to your github homepage
     # Uncomment below to add more social links
     # - https://www.facebook.com/username
     # - https://www.linkedin.com/in/username
@@ -47,7 +47,7 @@ google_site_verification: google_meta_tag_verification # change to your verifica
 baseurl: ''
 
 # Change to your timezone › http://www.timezoneconverter.com/cgi-bin/findzone/findzone
-timezone: Asia/Shanghai
+timezone: Europe/Madrid
 
 google_analytics:
   id: ''          # Fill with your Google Analytics ID
@@ -61,8 +61,8 @@ google_analytics:
     cache: false  # pv data local cache, good for the users from GFW area.
 
 disqus:
-  comments: false  # boolean type, the global switch for posts comments.
-  shortname: ''    # Fill with your Disqus shortname. › https://help.disqus.com/en/articles/1717111-what-s-a-shortname
+  comments: true  # boolean type, the global switch for posts comments.
+  shortname: 'learningalgorithms'    # Fill with your Disqus shortname. › https://help.disqus.com/en/articles/1717111-what-s-a-shortname
 
 
 # Prefer color scheme setting.
@@ -77,7 +77,7 @@ disqus:
 #
 #     dark   - Use the dark color scheme
 #
-theme_mode:   # [light|dark]
+theme_mode: dark  # [light|dark]
 
 # The CDN endpoint for images.
 # Notice that once it is assigned, the CDN url

_posts/2020-11-02-Index.md

@@ -0,0 +1,14 @@
+---
+title: Index
+date: 2020-11-04 12:42:00 +0100
+categories: []
+tags: []     # TAG names should always be lowercase
+pin: true
+---
+
+The purpose of this post is to serve as an **Index**.
+
+# Machine Learning
+1. [Brief introduction]({%post_url _posts/machine-learning/2020-11-05-introduction%})
+
+# Mathematics

_posts/machine-learning/2020-11-03-introduction.md

@@ -0,0 +1,102 @@
+---
+title: Brief introduction
+date: 2020-11-05 12:42:00 +0100
+categories: [Machine Learning,General]
+tags: [introduction]     # TAG names should always be lowercase
+image: /assets/img/img_posts/machine-learning/introduction/subsets_AI.png
+comments: true
+---
+
+## What is Artificial Intelligence?
+
+We can define **Artificial intelligence** (AI) as intelligence demonstrated by machines, contrasting with **natural intelligence** displayed by humans and animals, being intelligence described in a general form as the ability to discern information and use it to create knowledge that latter can be applied in similar situations. In this case, research in AI has been focused on learning, problem-solving, reasoning, perception, and use of language. 
+
+Talking about more concrete tasks, since the 1940s computers have been able to perform a complex task such as playing chess or proving mathematical theorems, problems that when solved by a machine have lost the property of being problems only solvables using intelligence, and therefore a task that doesn't need AI, the [AI effect](https://en.wikipedia.org/wiki/AI_effect).
+
+This era of research in AI began in 1935 when the mathematician [Alan Turing](https://en.wikipedia.org/wiki/Alan_Turing) proposed his general-purpose computer, the [Turing Machine](https://en.wikipedia.org/wiki/Turing_machine). In a [public lecture](https://www.vordenker.de/downloads/turing-vorlesung.pdf) in 1947 Turing said:
+
+>What we want is a machine that can learn from experience. The possibility of letting the machine alter its own instructions provides the mechanism for this.
+>
+> -- <cite>[Alan Mathison Turing](https://en.wikipedia.org/wiki/Alan_Turing)</cite>
+
+We could call this an early definition of **Machine Learning**. In a non-published paper in 1948 Turing also introduced an [idea](http://www.alanturing.net/turing_archive/pages/Reference%20Articles/connectionism/Turing's%20neural%20networks.html) to train a neural network of artificial neurons to perform a specific task, like in **Deep Learning**.
+
+## What is Machine Learning?
+
+The computer science discipline of **Machine Learning** (ML) is a subfield of **Artificial Intelligence**. ML objective is to construct algorithms that using a collection of examples of some phenomenon can improve their performance by itself.
+
+Let's see some examples of how experts have defined ML as:
+
+> Field of study that gives computers the ability to learn without being explicitly programmed.
+>
+> -- <cite>[Arthur L. Samuel](https://en.wikipedia.org/wiki/Arthur_Samuel)</cite>
+
+Arthur Samuel popularized the term _Machine Learning_ in 1959, and his Checkers-playing program was one of the first successful self-learning algorithms.
+
+> A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.
+>
+> -- <cite>[Tom M. Mitchell](https://en.wikipedia.org/wiki/Tom_M._Mitchell)</cite>
+
+Tom Mitchell is a former Chair of the Machine Learning Department at CMU. Mitchell is known for his contributions to the advancement of machine learning, artificial intelligence, and cognitive neuroscience.
+
+In summary, Machine Learning algorithms learn by experience, in a similar way that natural organisms do.  
+
+## What is Deep Learning?
+
+**Deep Learning** is a subfield of ML, and its methods are based on [artificial neural networks](https://en.wikipedia.org/wiki/Artificial_neural_network), which are algorithms inspired by the structure and function of the neurons, trying to simulate a brain. The term Deep Learning was introduced to the ML community by [Rita Dechter](https://en.wikipedia.org/wiki/Rina_Dechter) in 1986.
+
+> Deep Learning is constructing networks of parameterized functional modules & training them from examples using gradient-based optimization.
+>
+> -- <cite>[Yann LeCun](https://en.wikipedia.org/wiki/Yann_LeCun)</cite>
+
+In the last decade algorithms based on Deep Learning have produced results comparable to, and in some cases surpassing, human expert performance. Fields as diverse as: 
+
+- Computer vision:
+    - [Posenet: Pose detection](https://github.com/tensorflow/tfjs-models/tree/master/posenet)
+    - [YOLO: Real-Time Object Detection](https://pjreddie.com/darknet/yolo/)
+    - [3D Model From 2D Images](https://ai.facebook.com/blog/powered-by-ai-turning-any-2d-photo-into-3d-using-convolutional-neural-nets/)
+    - [Computer Vision in Healthcare Applications](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5857319/)
+    - [Self-driving cars](https://www.tesla.com/autopilot)
+
+- Speech recognition:
+    - [Google’s Google Assistant](https://assistant.google.com/intl/es_es/)
+    - [Speech recognition in healthcare](https://codeit.us/blog/speech-recognition-in-healthcare)
+    - [Speech recognition with Amazon's Alexa and the internet of things](https://docs.aws.amazon.com/es_es/iot/latest/developerguide/avs-integration-aws-iot.html)
+
+- Natural language processing:
+    - [Spam detector in emails](https://www.spamtitan.com/spam-email-filter/)
+    - [Search results with Google](https://blog.google/products/search/search-language-understanding-bert/)
+    - [DeepL Language translator](https://www.deepl.com/home)
+    - [GPT-3 Text generation](https://openai.com/blog/openai-api/)
+
+- Games:
+    - [AlphaZero in Go, Chess, and Shogi](https://en.wikipedia.org/wiki/AlphaZero)
+    - [AlphaStar in StarCraft II](https://deepmind.com/blog/article/AlphaStar-Grandmaster-level-in-StarCraft-II-using-multi-agent-reinforcement-learning)
+
+There are many examples of successful use of DL in today's world, and these models will only get better with more [data](https://www.forbes.com/sites/bernardmarr/2018/05/21/how-much-data-do-we-create-every-day-the-mind-blowing-stats-everyone-should-read/?sh=2440c20060ba).
+
+## How is performed the "Learning"?
+
+We can classify the type of learning used by the algorithm into four categories.
+
+### Supervised Learning
+
+In **supervised learning**, the dataset, the examples of the problem that we already have, are labeled and with them, our algorithms are going to infer a function, a **model** that would take inputs of the form of our examples and deduce an output. For example, a supervised learning algorithm could take as an input the square meters of a house and generate an output of how much it should cost.
+
+### Unsupervised Learning
+
+In **unsupervised learning**, our algorithm is going to try to find undetected patterns in the dataset, a dataset that doesn't contain any label.
+
+### Semi-supervised Learning
+
+**Semi-supervised** algorithms in machine-learning combine usually a small number of labeled examples and a much bigger quantity of unlabeled, the algorithm is going to use those unlabeled examples to produce a better model by being less biased. 
+
+
+### Reinforcement Learning
+
+In **reinforced learning**, our algorithm performs inside an environment in which is capable of perceiving its state and take actions, in consequence, those actions are going to report a benefit or a detriment, and this agent has to *learn* how to maximize its reward.
+
+And now, a little joke from [XKCD](https://xkcd.com/1838/)
+
+![XKCD Machine Learning](/assets/img/img_posts/machine-learning/introduction/xkcd-machine-learning.png)
+*XKCD Machine Learning*

_posts/mathematics/2020-11-16-linear-algebra.md

@@ -0,0 +1,55 @@
+---
+title: Basic Linear Algebra
+date: 2020-11-16 12:42:00 +0100
+categories: [Mathematics,Linear Algebra]
+tags: [algebra]     # TAG names should always be lowercase
+comments: true
+math: true
+---
+
+> You cannot know too much Linear Algebra
+>
+> -- <cite>[Benedict Hyman Gross](https://boffosocko.com/2012/05/05/you-cannot-learn-too-much-linear-algebra/)</cite>
+
+Prepostscript (PPS):  This post will be updated when more knowledge of linear algebra to understand other posts is needed.  
+
+## Introduction:
+
+You may have noticed that **Linear Algebra** is a term composed by _Linear_ + _Algebra_. So let's start by defining what is each of the two terms. An _algebra_ it's the union of a **set of objects** and a **set of rules** that applies to that objects. Etymologically the word _Algebra_ comes from the title of the book  [_Ilm al-jabr wa l-muqābala_](https://en.wikipedia.org/wiki/The_Compendious_Book_on_Calculation_by_Completion_and_Balancing), '_The Compendious Book on Calculation by Completion and Balancing_', by the Persian mathematician [al-Khwarizmi](https://en.wikipedia.org/wiki/Muhammad_ibn_Musa_al-Khwarizmi), also, the word _Algorithm_ derivates from his name, so you can imagine that he was an influential mathematician. The word _Linear_  refers to that what we are traying to solve using those set of objects and rules are **Linear systems**. Systems of equations where those equations are of first order, the maximum degree of the variables is equal to one, so in $$\mathbb R^2$$ (in the plane) we have straight lines, in $$ \mathbb R ^3$$​  (in a space) we have a plane.
+
+Those objects in our case, in ML, will be **vectors**, elements of $$\mathbb R ^n$$ (tuples of $$n$$ real numbers). For example, $$v$$ can be a vector of $$\mathbb R ^2$$ ($$ v \in \mathbb R ^2$$) with coordinates $$v_1 = 42$$ and $$v_2 = 73$$: 
+
+$$ v  =  \begin{bmatrix} 42 \cr 73 \end{bmatrix}\in \mathbb R ^2 $$
+
+We can represent a vector of $$w \in \mathbb R^n$$:
+
+$$ w  =  \begin{bmatrix} w_1 \cr w_2 \cr . \cr . \cr . \cr w_n \end{bmatrix}\in \mathbb R ^n $$
+
+And the **rules** or _axioms_ that apply to those vectors are the following:
+
+Being $$u, v, w$$ vectors, and $$a, b, 0$$ and $$1$$ _scalars_ (elements of $$\mathbb R$$), then:
+
+- (A1) $$ u + (v + w) = (u + v) + w$$
+- (A2) $$u + v = v + u$$
+- (A3) Exists a vector, called the zero vector, such that $$v +0 = v$$
+- (A4) For every vector $$v$$ exists a vector $$-v$$ called the inverse
+- (A5) $$a(u+v) = au + av$$
+- (A6) $$(a + b) v = av + bv$$
+- (A7) $$ a(bv)  = (ab)v$$
+- (A8) Exist a scalar called the identity, and denoted as $$1$$, such as $$1*v = v$$
+
+Mathematically those rules are important, since they are defining how vectors interact with themselves and with scalars. 
+
+Now, specifically, we can define those operations on vectors of $$\mathbb R^n$$ .  For example, if $$u, v \in \mathbb R^2$$ and $$a \in \mathbb R$$, then:
+
+$$ u + v = \begin{bmatrix} u_1 \cr u_2 \end{bmatrix} + \begin{bmatrix} v_1 \cr v_2 \end{bmatrix} = \begin{bmatrix} u_1 + v_1 \cr u_2+ v_2 \end{bmatrix} $$
+
+$$ a * u = a * \begin{bmatrix} u_1 \cr u_2 \end{bmatrix} = \begin{bmatrix} a * u_1 \cr a * u_2 \end{bmatrix}$$
+
+
+
+**References**:
+
+- [Mathematics for machine learning](https://mml-book.github.io/) by Marc Peter Deisenroth, Aldo Faisal, and Cheng Soon Ong
+- 3blue1brown's Linear Algebra [Videos](https://www.youtube.com/playlist?list=PLZHQObOWTQDPD3MizzM2xVFitgF8hE_ab)
+