Short-Term Forecasting of Stock Prices Using Long Short Term Memory

Kumar, Saurav; Ningombam, Dhruba

doi:10.1109/icit.2018.00046

Cited by 24 publications

(16 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Kumar and Ningombam perform long short-term memory (LSTM) to implement technical analysis for stock price forecasting of APPL in 2018 [16]. Their training details include dropout of 0.25, learning rate is 0.001, adam optimizer, epochs of 50, and linear activation for output layer activation.…”

Section: Unidirectional Long Short-term Memory (Denoted Model_2a)mentioning

confidence: 99%

Combining Time Series and Sentiment Analysis for Stock Market Forecasting

Chou¹,

Ramachandran²

2021

Proceedings of the 3rd International Conference on Statistics: Theory and Applications

View full text Add to dashboard Cite

Objective of this research is to build a model to predict stock price using sentimental information from news headlines and historical prices, and the model is able to not only conclude better results but also minimize the difference between predicted values and actual values. News headlines show impact on stock price. Unlike previous approaches where the textual information were usually calculated into sentiment score, we apply various approaches to extract information from news headlines. On the other hand, price data through time series are also useful to predict stock prices. Hence, improvement is made with combination of sentiment analysis of news headlines and time series analysis of historical prices, and the combination is able to complement nonavailability of sentiment lexicon and lack of news. Compared to time series models and word embedding models, our combined model shows smaller or similarly error measures MAPE, MAE, and RMSE with time series models, and reduces lag in graphs.

show abstract

Section: Unidirectional Long Short-term Memory (Denoted Model_2a)mentioning

confidence: 99%

Combining Time Series and Sentiment Analysis for Stock Market Forecasting

Chou¹,

Ramachandran²

2021

Proceedings of the 3rd International Conference on Statistics: Theory and Applications

View full text Add to dashboard Cite

show abstract

“…Borsa verilerinden asıl verimi elde etmek için kısa süreli tahmin yapmak özellikle dinamik kullanıcılar için önem arz etmektedir. Özellikle bir gün içerisindeki "Al-Sat" komutlarıyla kazanç sağlamak isteyen yatırımcılar farklı makine öğrenmesi algoritmaları kullanarak bu amaçlarına ulaşmaya çalışmışlardır [15,16]. Ancak bu tip sınıflandırma yapma işlemlerinde yeterli performans göstermekte zorluklarla karşılaşılması kaçınılmazdır.…”

Section: Literatürdeki çAlışmalarunclassified

“…Bu çalışmadaki veriler 2 Ocak 1968'den 9 Nisan 2018'e kadar olan NYSE, NASDAQ ve NYSE MKT'de işlem gören ABD merkezli hisse senetlerinin tamamı için günlük fiyat verilerinden oluşmaktadır. Veri girişi olarak, geliştirmiş olduğum 3 mimaride de son 5 günlük değerleri içeren 15 Bu çalışmada sunulan verilerin değerlendirilmesi, bunun derin öğrenme algoritmalarıyla yeterince doğru bir çözüm bulmak için yeterli olduğunu göstermektedir. Bu veri setinde, Tablo 1'de gösterildiği gibi veri setimizde, sistemimiz üzerinde çalışacak 5 özelliğe sahibiz.…”

Section: Veri Setiunclassified

Derin Öğrenme Yöntemleri ile Borsada Fiyat Tahmini

Koçer

Önde

Şahingöz

2020

Bitlis Eren Üniversitesi Fen Bilimleri Dergisi

View full text Add to dashboard Cite

ÖzSon yıllarda, bilgisayarların donanımındaki teknolojik gelişmeler ve makine öğrenme tekniklerindeki gelişmeler nedeniyle, "Büyük Veri" ve "Paralel İşleme" kullanımı olmak üzere problem çözmek için iki artan yaklaşım vardır. Özellikle GPU'lar gibi çok çekirdekli bilgi işlem aygıtlarında paralel olarak gerçekleştirilebilen Derin Öğrenme algoritmalarının ortaya çıkmasıyla, bu yaklaşımlarla birçok gerçek dünya problemleri çözülebilmektedir. Derin öğrenme modelleri eğitildikleri veri ile sınıflandırma, regresyon analizi ve zaman serilerinde tahmin gibi uygulamalarda büyük başarılar göstermektedir. Bu modellerin finansal piyasadaki en aktif uygulama alanlarından biri özellikle borsada işlem gören hisse senetlerinin tahmini işlemleridir. Bu alanda amaç, pazardaki değişim süreci hakkındaki hisse senedinin önceki günlük verilerine bakarak kısa veya uzun vadeli gelecekteki değerini tahmin etmeye çalışmaktır. Bu çalışmada, LSTM, GRU ve BLSTM isimli 3 farklı derin öğrenme modeli kullanılarak bir hisse senedi tahmin sistemi geliştirilip, kullanılan modeller arasında karşılaştırmalı bir analiz yapıldı. Spekülatif hareketlerden uzak olması için veri seti olarak 1968'den 2018'e kadar olan New York Borsası'ndan hisse senedinin zaman serisi değerleri kullanıldı. Spesifik olarakta IBM hisse senedi ile test çalışmaları yapıldı. Deneysel sonuçlar, hisse senedine ait son 5 günlük işlem verilerinin girdi olarak kullanıldığı BLSTM modeliyle yapılan eğitimin %63,54 lük bir yönsel doğruluk değerine ulaşıldığını göstermektedir. AbstractIn recent years, due to the technological advances in computer hardware and enhancements in machine learning techniques, there are two hot research areas in problem solving, the use of "Big Data" and "Parallel Processing". Many real-world problems can be solved with the use of different Deep Learning algorithms, which can be realized in parallel with multicore computing devices such as GPUs. Deep learning models show great success in applications such as classification of raw data, regression analysis and estimation in time series. One of the most active application areas of these models is "the financial market" which aims a good estimation of stock prices in the exchange market. In this paper, it is aimed to forecast the short or long term future value by looking at the previous log data of the stock on the process of change in the market. In this study, a price forecasting system was developed by using 3 different deep learning models named LSTM, GRU and BLSTM with a comparative analysis between them. The time series values of the stock were used from the New York Stock Exchange from 1968 to 2018 as a set of data to be free of speculative movements. Specifically, tests were conducted on IBM stock. Experimental results show that the directional accuracy of 63.54% was achieved with the BLSTM model where the last 5-day transaction data of the stock were used as input.

show abstract

“…Recently, many deep learning approaches are used to predict the chaotic time series, which is called the fourth category CTSPM in this paper. Like long short term memory (LSTM) neural network and its improved models [31], [32]. Those CTSPMs make good progress and provide references for the prediction of the KHT time series in this paper.…”

Section: Introductionmentioning

confidence: 99%

An Accurate Online Prediction Model for Kiln Head Temperature Chaotic Time Series

Zhang

Chen

et al. 2020

IEEE Access

View full text Add to dashboard Cite

The kiln head temperature (KHT) is a key thermal variable to condition monitoring and stable control of the rotary kiln. In this paper, a hybrid dynamic prediction framework (HDPF) is proposed to realize the accurate online prediction of the KHT chaotic time series of the rotary kiln. Where, in order to response time-delay and time-varying behaviors of the rotary kiln, the phase space reconstruction (PSR) method is used to achieve the high-dimensional dynamic trajectory of the sintering system of the rotary kiln and the local Volterra filter based on the sliding window technique is developed to construct the dynamic relationship from the high-dimensional dynamic trajectory. Furthermore, considering the practical industrial data contains unreliable risk sometimes, each predicted value is corrected by the offset compensation technique as the final output. Based on well-known Mackey-Glass chaotic time series and real-world KHT chaotic time series from the Inner Mongolia branch of Datang International Power Generation Corporation Limited, comparative experiments show that the HDPF has the highest prediction accuracy which means the HDPF can capture the dynamic evolution of the KHT over time and has the strongest robustness for four kinds of noise which means it can be applied in practical industries with noise. Meanwhile, it has low computational complexity which can meet the real-time requirements of the monitor and control. Further, those good performances imply the HDPF is effective not only for the rotary kiln but also for other complex nonlinear systems with the time-delay and time-varying behaviors. INDEX TERMS Rotary kiln, kiln head temperature, chaotic time series prediction, Volterra filter, offset compensation.

show abstract

Short-Term Forecasting of Stock Prices Using Long Short Term Memory

Cited by 24 publications

References 4 publications

Combining Time Series and Sentiment Analysis for Stock Market Forecasting

Combining Time Series and Sentiment Analysis for Stock Market Forecasting

Derin Öğrenme Yöntemleri ile Borsada Fiyat Tahmini

An Accurate Online Prediction Model for Kiln Head Temperature Chaotic Time Series

Contact Info

Product

Resources

About