Predicting the stock price trends using a K-nearest neighbors-probabilistic model

Lock Siew Han, Md. Jan Nordin

Research output: Contribution to journalArticle

Abstract

This paper examines a hybrid model which combines a K-Nearest Neighbors (KNN) approach with a probabilistic method for the prediction of stock price trends. One of the main problems of KNN classification is the assumptions implied by distance functions. The assumptions focus on the nearest neighbors which are at the centroid of data points for test instances. This approach excludes the non-centric data points which can be statistically significant in the problem of predicting the stock price trends. For this it is necessary to construct an enhanced model that integrates KNN with a probabilistic method which utilizes both centric and non-centric data points in the computations of probabilities for the target instances. The embedded probabilistic method is derived from Bayes’ theorem. The prediction outcome is based on a joint probability where the likelihood of the event of the nearest neighbors and the event of prior probability occurring together and at the same point in time where they are calculated. The proposed hybrid KNN-Probabilistic model was compared with the standard classifiers that include KNN, Naive Bayes, One Rule (OneR) and Zero Rule (ZeroR). The test results showed that the proposed model outperformed the standard classifiers which were used for the comparisons.

Original languageEnglish
Pages (from-to)6245-6255
Number of pages11
JournalJournal of Theoretical and Applied Information Technology
Volume96
Issue number18
Publication statusPublished - 30 Sep 2018

Fingerprint

Stock Prices
Probabilistic Model
Nearest Neighbor
Classifiers
Probabilistic Methods
Classifier
Bayes' Formula
Prior Probability
Naive Bayes
Prediction
Hybrid Model
Distance Function
Centroid
Trends
Statistical Models
Likelihood
Integrate
Target
Necessary
Zero

Keywords

  • Bayes’ theorem
  • K-nearest neighbors
  • Naive bayes
  • Probabilistic method
  • Stock price prediction

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Predicting the stock price trends using a K-nearest neighbors-probabilistic model. / Han, Lock Siew; Nordin, Md. Jan.

In: Journal of Theoretical and Applied Information Technology, Vol. 96, No. 18, 30.09.2018, p. 6245-6255.

Research output: Contribution to journalArticle

@article{0d72046f0d6341f4802cd509b2d92e49,
title = "Predicting the stock price trends using a K-nearest neighbors-probabilistic model",
abstract = "This paper examines a hybrid model which combines a K-Nearest Neighbors (KNN) approach with a probabilistic method for the prediction of stock price trends. One of the main problems of KNN classification is the assumptions implied by distance functions. The assumptions focus on the nearest neighbors which are at the centroid of data points for test instances. This approach excludes the non-centric data points which can be statistically significant in the problem of predicting the stock price trends. For this it is necessary to construct an enhanced model that integrates KNN with a probabilistic method which utilizes both centric and non-centric data points in the computations of probabilities for the target instances. The embedded probabilistic method is derived from Bayes’ theorem. The prediction outcome is based on a joint probability where the likelihood of the event of the nearest neighbors and the event of prior probability occurring together and at the same point in time where they are calculated. The proposed hybrid KNN-Probabilistic model was compared with the standard classifiers that include KNN, Naive Bayes, One Rule (OneR) and Zero Rule (ZeroR). The test results showed that the proposed model outperformed the standard classifiers which were used for the comparisons.",
keywords = "Bayes’ theorem, K-nearest neighbors, Naive bayes, Probabilistic method, Stock price prediction",
author = "Han, {Lock Siew} and Nordin, {Md. Jan}",
year = "2018",
month = "9",
day = "30",
language = "English",
volume = "96",
pages = "6245--6255",
journal = "Journal of Theoretical and Applied Information Technology",
issn = "1992-8645",
publisher = "Asian Research Publishing Network (ARPN)",
number = "18",

}

TY - JOUR

T1 - Predicting the stock price trends using a K-nearest neighbors-probabilistic model

AU - Han, Lock Siew

AU - Nordin, Md. Jan

PY - 2018/9/30

Y1 - 2018/9/30

N2 - This paper examines a hybrid model which combines a K-Nearest Neighbors (KNN) approach with a probabilistic method for the prediction of stock price trends. One of the main problems of KNN classification is the assumptions implied by distance functions. The assumptions focus on the nearest neighbors which are at the centroid of data points for test instances. This approach excludes the non-centric data points which can be statistically significant in the problem of predicting the stock price trends. For this it is necessary to construct an enhanced model that integrates KNN with a probabilistic method which utilizes both centric and non-centric data points in the computations of probabilities for the target instances. The embedded probabilistic method is derived from Bayes’ theorem. The prediction outcome is based on a joint probability where the likelihood of the event of the nearest neighbors and the event of prior probability occurring together and at the same point in time where they are calculated. The proposed hybrid KNN-Probabilistic model was compared with the standard classifiers that include KNN, Naive Bayes, One Rule (OneR) and Zero Rule (ZeroR). The test results showed that the proposed model outperformed the standard classifiers which were used for the comparisons.

AB - This paper examines a hybrid model which combines a K-Nearest Neighbors (KNN) approach with a probabilistic method for the prediction of stock price trends. One of the main problems of KNN classification is the assumptions implied by distance functions. The assumptions focus on the nearest neighbors which are at the centroid of data points for test instances. This approach excludes the non-centric data points which can be statistically significant in the problem of predicting the stock price trends. For this it is necessary to construct an enhanced model that integrates KNN with a probabilistic method which utilizes both centric and non-centric data points in the computations of probabilities for the target instances. The embedded probabilistic method is derived from Bayes’ theorem. The prediction outcome is based on a joint probability where the likelihood of the event of the nearest neighbors and the event of prior probability occurring together and at the same point in time where they are calculated. The proposed hybrid KNN-Probabilistic model was compared with the standard classifiers that include KNN, Naive Bayes, One Rule (OneR) and Zero Rule (ZeroR). The test results showed that the proposed model outperformed the standard classifiers which were used for the comparisons.

KW - Bayes’ theorem

KW - K-nearest neighbors

KW - Naive bayes

KW - Probabilistic method

KW - Stock price prediction

UR - http://www.scopus.com/inward/record.url?scp=85055275670&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85055275670&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:85055275670

VL - 96

SP - 6245

EP - 6255

JO - Journal of Theoretical and Applied Information Technology

JF - Journal of Theoretical and Applied Information Technology

SN - 1992-8645

IS - 18

ER -