Machine Learning - Future Artificial Intelligence

---

Artificial intelligence will shape our future more powerfully than any other innovation of this century. Anyone who doesn't understand it will soon find that he feels forgotten and wakes up in a world full of technology that feels more and more like magic.
In 2015, Google trained a conversational agent (AI) that not only convincingly interacts with humans as a technical support desk, but also discusses ethics, expressing opinions and answering general questions based on facts.

---

Machine learning + artificial intelligence

---

Machine learning is one of many sub-areas of artificial intelligence, involving the way computers learn from experience to improve their ability to think, plan, decide, and act

Its goal is to enable computers to learn on their own. The machine's learning algorithm enables it to identify patterns in the observed data, build models that explain the world, and predict things without explicit pre-programmed rules and models.

---

Machine learning workflow

---

The machine learning workflow is the process required to execute a machine learning project. While individual projects may vary, most workflows share several common tasks: problem assessment, data exploration, data preprocessing, model training/testing/deployment, and more. You will find useful visualizations of these core steps below:

---

Machine learning application

---

As already mentioned, machine learning is a technology that helps develop and improve many of the features we are used to. Let's consider the ones we most often encounter when we use our favorite applications on a daily basis, and finally understand the advantages of machine learning technology.

---

Gesture Recognition

---

They use algorithms through acoustic and language modeling. Acoustic modeling represents the connection between language units of speech and audio signals, and language modeling matches sounds to word sequences to distinguish words that sound similar.

This allows users to talk to their computers and convert their text into text through word processing and speech recognition.

You can access feature commands such as setting an alarm, opening a file, booking at your favorite restaurant, and more. On the other hand, some mobile applications are used for precise business settings, such as medical or legal transcription.

---

Image Identification

---

This feature is very common in mobile apps. For example, it is sometimes used for identification purposes or to use photos that include filters and edits. In addition, using different types of machine learning algorithms, you can define the user's gender and age in the application to achieve eye retina or fingerprint recognition. A good example of a machine learning application is the identification of license plate violations on the road on the road.

---

Intelligent data analysis

---

The more you collect and process user data in tandem big data and machine learning, the more you know the features they often use or not use at all. This way, you can expand your understanding of your audience and adapt your app to your liking to make your service even better.

This combination has been applied by Amazon and Google in some of their services.

---

Python or R

---

The R language is very popular among data scientists, but if you need regular programming, it has many drawbacks. On the other hand, Python is a general-purpose programming language that can be applied to many use cases.

This may be the main reason why Python has become a machine and deep learning field in recent years. Each decent library provides the Python API or treats it as the only target language.

Python is a very simple and beginner-friendly language. Moreover, there is no need to know all the complexity of the language to apply it to ML.

---

Jupyter Notebook

---

In traditional programming, most of the time is spent in a text editor or IDE, and in data science, most of the code is written in Jupyter Notebook.

It is a simple and powerful tool for patching data analysis problems. It allows you to write Python code, text descriptions, and embed charts and charts directly into interactive web pages.

To make things easier, Google created a free service, Google Colab, which provides CPU resources and access to the GPU unit, which is handy when dealing with neural networks and deep learning.

---

Scikit-learn

---

Scikit-learn is one of the most popular ML libraries today. It supports most classic supervised and unsupervised learning algorithms: linear and logistic regression, SVM, naive Bayes, gradient enhancement, clustering, k-means, and more.

In addition to different ML models, Scikit-learn provides a variety of data preprocessing and results analysis methods. Scikit-learn focuses on the classical ML algorithm, so its support for neural networks is very limited and cannot be used for deep learning problems.

---

PyTorch

---

PyTorch is a popular deep learning library built by Facebook. In addition to the CPU, it also supports GPU accelerated computing. The library is dedicated to providing users with a fast and flexible modeling experience and has gained great appeal in the deep learning community.

---

Matplotlib

---

Matplotlib is the standard tool in every data scientist toolbox. It provides the ability to draw many different types of graphs and charts to display results.

Matplotlib diagrams can be easily embedded in Jupyter Notebook. This way, you can always visualize the data and results you get from the model.

---

Program Framework Summary

---

Despite a large ecosystem, the most powerful tools in the world of machine learning and artificial intelligence are the most widely used tools. When you're just getting into a machine learning journey, it's a good idea to choose from these tools as they are widely introduced in a variety of tutorials, stepping-out courses, and have good community support.

To simplify the learning complexity, you can start with the classic ML algorithm and pay more attention to the use of Scikit-learn. After learning more about it, continue to use Tensorflow, PyTorch or Keras and enter the world of deep learning.