An algorithm is a set of well-defined, ordered and finite instructions or rules that allow you to carry out an activity through successive steps. There is an initial state and an input, and then the successive steps will reach a final state and a solution is obtained. In the “real” world, a step-by-step instructions book would be an algorithm.
In the digital world, algorithms have become a fundamental part of our lives since computers are very good at executing these steps with precision and speed. A well-defined algorithm can have total autonomy and offer a differential value to the final user. The great change of paradigm that the algorithms have lived has been thanks to Artificial Intelligence and its new ways of learning. Algorithms used to be based on expert systems with the knowledge incorporated in them as a set of rules, but with the arrival of more complex and changing contexts, the models had to be alive and supported by much more advanced learning systems. This is where machine and deep learning comes in, incorporating knowledge into the parameters of a cognitive model.
Do you know that every day you interact with systems managed by algorithms? Here are some of them:
- Google Search. The famous PageRank that organizes the results and in which the SEO (and part of the SEM) is based
- Facebook news. The also known EdgeRank, as controversial for brands as effective for users.
- OKCupid Date Matching. Algorithm that uses this dating platform that can brag is the most advanced and effective in its category.
- Amazon “You can also enjoy …”. Reknown algorithm of products that appear in all Amazon pages.
- MP3 compression. Algorithm that compresses information (in this case musical but in video is the same), reducing the file size with the less quality loss possible.
- Auto-Tune. Did not you know that auto-tune is an algorithm? It is not as sophisticated as the others but it processes sound by a series of rules and modifies the tone adjusting it to where it is wanted.
Now with the Internet of Things (IoT) everything will be managed by algorithms, from driving autonomous cars to smart home devices (light, heating, vacuum …). In the case of autonomous cars, as it happens in projects of facial recognition, artificial vision is used where the whole process is carried out using an algorithm.
If you want to get a little more technical, I’m also sharing some interesting links:
- 7 Algorithms That Rule The World. YouTube video that reviews the algorithms from the mathematical / statistical point of view. For example: Fast Fourier Transfusion, Dijkstra Algorithm or RSA Algorithm.
- Modern Machine Learning Algorithms: strengths and weaknesses. They explain in detail the main algorithms of Machine Learning: Regression, Classification and Clustering.
- The 10 Algorithms Machine Learning Engineers Need to Know. Another great article based on Machine Learning algorithms, in this case less mathematical, but technological.
- Beyond Backpropagation: can we go deeper than deep learning? An excellent article, in this case focused on Deep Learning and the algorithms and techniques used for their models.
#365daysof #futurism #innovation #technology #algorithms #day46
Source: Deep Learning on Medium