Xxii республикалық студенттер мен жас ғалымдардың ғылыми конференция материалдары


OVERVIEW OF STATISTICAL LEARNING METHODS IN THE FIELD OF



Pdf көрінісі
бет105/267
Дата18.10.2023
өлшемі8,13 Mb.
#117865
1   ...   101   102   103   104   105   106   107   108   ...   267
Байланысты:
Сборник материалов конференции (продолжение)

 
OVERVIEW OF STATISTICAL LEARNING METHODS IN THE FIELD OF 
PREDICTIVE MEDICINE 
Iglikov T.D., Atymtayeva L.B. 
International IT University (IITU) 
Abstract 
The whole humanity seeks for ideas in enhancing the healthcare and one of the most 
important goals of medicine is making the correct diagnoses. Diagnoses accuracy is highly 
important, because of possibility saving many human lives with a proper treating. Nowadays this 
problem can be solved not only from medicine prospective, but with computer power and 
statistical theory. So, in this article covered an overview of applicable methods of statistical 
learning, which could help in predicting different diseases using databases with patients’ medical 
indicators or any other useful and meaningful digital information. The main idea of this article is 
to examine different statistical learning techniques in the field of medicine for finding well 
interpretable and effectively predicting models, which could increase probability of correct and 
forehanded detection of diseases. Hopely, implementing such models may lead to higher quality 
of patient care in any medical organization. 
Key words:
statistical learning, disease prediction, computer science, supervised learning, 
unsupervised learning, learning problem, logistic regression, decision tree, random forest, k-
nearest neighbors, support vector machines 
At first it’s important to understand, what is statistical learning and machine learning? 
And what is difference between statistical learning and machine learning? For many people it 
may seem that it is the same collocations and even technical students have a vague 
understanding. Both statistical learning and machine learning are aiming on extracting some 
useful information from datasets, creating model and using it for predictions. Machine learning 
seeking for accurate predictions for unobserved data and statistical learning looking for strong 
inferences between observations and outcomes. In many practical cases it’s not important to 
have a clear inference, rather than accurate and precise predictions, for example for companies 
who want to predict future prices of cars, houses and etc. it’s not important how the prediction 
was made, their main focus is on accuracy of predictions, so that the company may get the profit 
of new data. In opposite for predicting different diseases in medicine very influential and critical 
to know how these predictions were made, how features correlated with outcome and why some 
of features are more significant than others. Many methods could be used either in statistical 
learning or in machine learning and all of them could be divided into two big categories: 
supervised learning and unsupervised learning. 
Supervised learning based on making predictions in situation when we already have 
outputs for our inputs. The key idea is to build efficient model, which will be capable to make 
decent predictions for future inputs. However in unsupervised learning we facing no outputs to 
our inputs and commonly It’s the case of clustering and grouping problem, when we trying to 
find some patterns or relations inside of the dataset[1]. For sure unsupervised learning is much 
more challenging and complicated challenge for solving. 


227 


Достарыңызбен бөлісу:
1   ...   101   102   103   104   105   106   107   108   ...   267




©emirsaba.org 2024
әкімшілігінің қараңыз

    Басты бет