The following file may be helpful. Best of luck.

Classification algorithms (supervised algorithms predicting categorical labels) 1-Linear discriminant analysis.

2-Quadratic discriminant analysis.

3-Maximum entropy classifier (aka logistic regression, multinational logistic regression): Note that logistic regression is an algorithm for classification, despite its name.

Dear H. Naderpour,

Pattern recognition methods:

1. Statistical pattern recognition

2. Neural networks

3. Structural pattern recognition

4. Syntactic pattern recognition

5. Approximate reasoning approach to pattern recognition

6. A logical combinatorial approach to pattern recognition

7. Applications of Support Vector Machine (SVM) for pattern recognition

and etc.

Regards, Shafagat

The following file may be helpful. Best of luck.

I agree with Shafagat .

It is a method of constructing multilayer neural networks based on information theory and genetic algorithms:

http://lc.kubagro.ru/aidos/_Aidos-X.htm

http://chaosandcorrelation.org/ForumCC/index.php

https://arxiv.org/pdf/1710.07264.pdf

There are many artificial intelligence systems. Universal cognitive analytical system "Eidos-x++" differs from them in the following parameters:

- developed in a universal setting, independent of the subject area. Therefore, it is universal and can be applied in many subject areas (http://lc.kubagro.ru/aidos/index.htm);

- is in full open free access (http://lc.kubagro.ru/aidos/_Aidos-X.htm), and with relevant source texts (http://lc.kubagro.ru/__AIDOS-X.txt);

- is one of the first domestic systems of artificial intelligence of the personal level, i.e. it does not require special training from the user in the field of artificial intelligence technologies (there is an act of introduction of the "Eidos" system of 1987) (http://lc.kubagro.ru/aidos/aidos02/PR-4.htm);

- provides stable detection in a comparable form of force and direction of cause-and-effect dependencies in incomplete noisy interdependent (nonlinear) data of very large dimension of numerical and non-numerical nature, measured in different types of scales (nominal, ordinal and numerical) and in different units of measurement (i.e. does not impose strict requirements on the data that can not be performed, and processes the data that is);

- contains a large number of local (supplied with the installation) and cloud-based educational and scientific applications (currently 31 and 133, respectively) (http://lc.kubagro.ru/aidos/Presentation_Aidos-online.pdf);

- provides multi-language interface support in 44 languages. Language databases are included in the installation and can be replenished automatically;

- supports on-line learning environment and is widely used all over the world (http://aidos.byethost5.com/map3.php);

- the most computationally intensive operations of model synthesis and recognition are implemented with the help of a graphics processor, which on some tasks provides acceleration of solving these problems by several thousand times, which really provides intelligent processing of large data, large information and large knowledge;

- provides conversion of initial empirical data into information, and its knowledge and solution using this knowledge of classification problems, decision support and research of the subject area by studying its system-cognitive model, while generating a very large number of tabular and graphical output forms (development of cognitive graphics), many of which have no analogues in other systems (examples of forms can be found in: http://lc.kubagro.ru/aidos/aidos18_LLS/aidos18_LLS.pdf)

You may follow the pdf of Reza Biria.

Pattern recognition techniques can in general be categorized into two classes: parametric and non-parametric. Choosing the best method depends on many factors such as the computational power, the amount of available data, dimension of the feature space, the distribution of your data, the application and the task.

If you provide more information about the problem, it would be easier to choose the best algorithm for that.

Regarding the best algorithm for classification, you are referred to the paper entitled "Classifier Technology and the Illusion of the Progress" which can be found in https://www.jstor.org/stable/27645729?seq=1#metadata_info_tab_contents

The best algorithm of pattern recognition depends on class of problems to which refers your problem.If conditional probabilities distributions of objects of different classes are known you may use Bayesian method of classification. If conditional pobabilities are not known you may use disc riminant methods one of the best methods of this class is SVM. If you need to recognize optical images one of the best approaches is Convolutional neural networks/

I suggest to use artificial neural network either multi-layer perceptron (MLP) or Radial Basis Functions networks (RBF). Both of these structures can be implemented using MATLAB. Python scikit package could be used also. Follow the steps (load data, pre-process data (input, output) sets, divide the data samples sets into (training, validation and testing sets), created ANN model, test the model using R-square to compare ANN output with the actual output.

I should comment on those answers suggesting to use a specific (and sophisticated) classifier such as neural networks or SVM.

It is not a good practice to blindly select a classifier without knowing anything about the task, the distribution of data, and the confounding factors in the available data set.

It is not always the case that a more complex and highly non-linear classifier (e.g. CNN) works better than a simpler one (such as a linear discriminant analysis).

I highly recommend you to read the very interesting paper by David Hand (I mentioned it in my previous answers). He showed that if one does not know much about confounding factors and sources of variability (which is more often the case), then it is better to use a simpler classifier instead of a more complex one as the large gains in classification accuracy can be achieved using relatively simple models. Moreover, the simple models are preferred since they are more interpretable than the sophisticated ones.

Here, I just quote one of the key sentences from that paper.

• "In particular, simple methods typically yield performance almost as good as more sophisticated methods, to the extent that the difference in performance may be swamped by other sources of uncertainty that generally are not considered in the classical supervised classification paradigm."

Following

Hello,

There is 2 main groups of pattern recognition techniques, traditional image processing and AI based applications, both techniques are evaluated based on many parameter, the market is full of research concerning those tools.

Regardless the technique, what it really matters is how patterns are defined, what is there scale, shape, location ... in order to define a good composition of an object.

Thanks,

There are no the best pattern recognition methods. You need to solve a concrete problem. To do this, you need to make its formalization, determine the information, the goals and the conditions, and so on. And only then among the acceptable methods can we choose the best methods. In reality, the situation is as follows. Having studied some methods in university, experts use them to solve any problem (provided they are valid). There is nothing insufficiency in this, if one does not forget about the inductive nature of the pattern recognition problem.

Good suggestion at all

Nowadays, deeplearning provides evidences to get the most reliability findings in pattern recognition.

@Ashwan deep neural networks can perhaps provide the most accurate results but not the most reliable ones. How a blackbox that is not interpretable can provide a reliable result?

The field of pattern recognition is concerned with the automatic discovery of regularities in data through the use of computer algorithms and with the use of these regularities to take actions such as classifying the data into different categories.