---

Machine Learning - Skills

---

---

---

Machine Learning experts need skills across several domains. Here is a list of the skills you'll need to acquire.

• Statistics

• Probability Theories

• Calculus

• Optimization techniques

• Visualization

Necessity of Various Skills of Machine Learning

You need the following skills:

Mathematical Notation

It is important that you know the notation used by mathematicians in their equations. Machine learning algorithms are heavily based on mathematics. It is probably just a beginner level level of mathematics you need to know. In order to learn machine learning, you need to be able to read the notation and comprehend what it means. Otherwise, you'll need to brush up on your math skills.

 

fANγnet−θ−γifnet−θ≥ϵif−ϵ<net−θ<ϵifnet−θ≤−ϵ$$f_{AN}(net-\theta )=\{\begin{array}{ll}\gamma & ifnet-\theta \ge ϵ\\ net-\theta & if-ϵ<net-\theta <ϵ\\ -\gamma & ifnet-\theta \le -ϵ\end{array}$$

$$f_{AN}(net-\theta)=\begin{cases}\gamma & if\:net-\theta \geq \epsilon\\net-\theta & if - \epsilon< net-\theta <\epsilon\\ -\gamma & if\:net-\theta\leq- \epsilon\end{cases}$$

 

maxα[∑i=1mα−12∑i,j=1mlabel(i)⋅label(j)⋅ai⋅aj⟨x(i),x(j)⟩]$$\begin{gathered} {\displaystyle } \\ \underset{\alpha }{max}\left[\begin{array}{c}{\displaystyle \sum _{i=1}^m\alpha -\frac{1}{2}{\displaystyle \sum _{i,j=1}^{m}labe{l}^{\left(\begin{array}{c}i\end{array}\right)}\cdot labe{l}^{\left(\begin{array}{c}j\end{array}\right)}\cdot a_i\cdot a_j⟨x^{\left(\begin{array}{c}i\end{array}\right)},x^{\left(\begin{array}{c}j\end{array}\right)}⟩}}\end{array}\right] \end{gathered}$$

$$\displaystyle\\\max\limits_{\alpha}\begin{bmatrix}\displaystyle\sum\limits_{i=1}^m \alpha-\frac{1}{2}\displaystyle\sum\limits_{i,j=1}^m label^\left(\begin{array}{c}i\\ \end{array}\right)\cdot\:label^\left(\begin{array}{c}j\\ \end{array}\right)\cdot\:a_{i}\cdot\:a_{j}\langle x^\left(\begin{array}{c}i\\ \end{array}\right),x^\left(\begin{array}{c}j\\ \end{array}\right)\rangle \end{bmatrix}$$

 

$$f_{AN}(net-\theta )=\left(\frac{e^{\lambda (net-\theta )}-e^{-\lambda (net-\theta )}}{e^{\lambda (net-\theta )}+e^{-\lambda (net-\theta )}}\right)$$

$$f_{AN}(net-\theta)=\left(\frac{e^{\lambda(net-\theta)}-e^{-\lambda(net-\theta)}}{e^{\lambda(net-\theta)}+e^{-\lambda(net-\theta)}}\right)\;$$

 

Probability Theory

Conditional probabilities are a great way to test your probability knowledge.

=p(x,y|ci)p(ci)p(x,y)$$p(c_i|x,y)=\frac{p(x,y|c_i)p(c_i)}{p(x,y)}$$

$$p(c_{i}|x,y)\;=\frac{p(x,y|c_{i})\;p(c_{i})\;}{p(x,y)\;}$$

 

By using these definitions, we can define the Bayesian classification rule−

• If P(c1|x, y) > P(c2|x, y) , the class is c1 .

• If P(c1|x, y) < P(c2|x, y) , the class is c2 .

Optimization Problem

maxα[∑i=1mα−12∑i,j=1mlabel(i)⋅label(j)⋅ai⋅aj⟨x(i),x(j)⟩]$$\begin{gathered} {\displaystyle } \\ \underset{\alpha }{max}\left[\begin{array}{c}{\displaystyle \sum _{i=1}^m\alpha -\frac{1}{2}{\displaystyle \sum _{i,j=1}^{m}labe{l}^{\left(\begin{array}{c}i\end{array}\right)}\cdot labe{l}^{\left(\begin{array}{c}j\end{array}\right)}\cdot a_i\cdot a_j⟨x^{\left(\begin{array}{c}i\end{array}\right)},x^{\left(\begin{array}{c}j\end{array}\right)}⟩}}\end{array}\right] \end{gathered}$$

$$\displaystyle\\\max\limits_{\alpha}\begin{bmatrix}\displaystyle\sum\limits_{i=1}^m \alpha-\frac{1}{2}\displaystyle\sum\limits_{i,j=1}^m label^\left(\begin{array}{c}i\\ \end{array}\right)\cdot\:label^\left(\begin{array}{c}j\\ \end{array}\right)\cdot\:a_{i}\cdot\:a_{j}\langle x^\left(\begin{array}{c}i\\ \end{array}\right),x^\left(\begin{array}{c}j\\ \end{array}\right)\rangle \end{bmatrix}$$

 

Here is an optimization function

Subject to the following constraints −

maxα[∑i=1mα−12∑i,j=1mlabel(i)⋅label(j)⋅ai⋅aj⟨x(i),x(j)⟩]$$\begin{gathered} {\displaystyle } \\ \underset{\alpha }{max}\left[\begin{array}{c}{\displaystyle \sum _{i=1}^m\alpha -\frac{1}{2}{\displaystyle \sum _{i,j=1}^{m}labe{l}^{\left(\begin{array}{c}i\end{array}\right)}\cdot labe{l}^{\left(\begin{array}{c}j\end{array}\right)}\cdot a_i\cdot a_j⟨x^{\left(\begin{array}{c}i\end{array}\right)},x^{\left(\begin{array}{c}j\end{array}\right)}⟩}}\end{array}\right] \end{gathered}$$

$$\displaystyle\\\max\limits_{\alpha}\begin{bmatrix}\displaystyle\sum\limits_{i=1}^m \alpha-\frac{1}{2}\displaystyle\sum\limits_{i,j=1}^m label^\left(\begin{array}{c}i\\ \end{array}\right)\cdot\:label^\left(\begin{array}{c}j\\ \end{array}\right)\cdot\:a_{i}\cdot\:a_{j}\langle x^\left(\begin{array}{c}i\\ \end{array}\right),x^\left(\begin{array}{c}j\\ \end{array}\right)\rangle \end{bmatrix}$$

 

All you need to do is read and understand the above.

Visualization

In order to interpret the algorithm's results, you must understand the various types of visualization plots.

Additionally to the above theoretical aspects of machine learning, good programming skills are required.

What it takes to implement machine learning will be discussed in the next chapter.

---

---

Frequently Asked Questions

---

---

---

---

---