Predicting Stock Prices Using Machine Learning Techniques


Karthikeyan C., YILDIZ E., Anandan P., Prabha R., Mohan D., Babu V. D.

6th International Conference on Inventive Computation Technologies (ICICT), Coimbatore, Hindistan, 20 - 22 Ocak 2021, ss.1184-1188 identifier identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/icict50816.2021.9358537
  • Basıldığı Şehir: Coimbatore
  • Basıldığı Ülke: Hindistan
  • Sayfa Sayıları: ss.1184-1188
  • Anahtar Kelimeler: Root Mean Square Error, Moving Average, Linear Regression
  • İstanbul Üniversitesi Adresli: Evet

Özet

Stock value examination to a great extent relies upon the capacity to recognize the development of the stock costs and foresee the concealed examples and patterns which the market follows. Information examinations have been developing significance on the financial exchange in the ongoing years. To get the benefit of the contributing, numerous financial specialists need to realize how to examine the significant information from the securities exchange. In a lot of general writing on stock foresee, it a couple of explicit direction show up on the future forecast. Thusly, how to anticipate the stocks from the recovery information, it turns into a significant and impressive issue on market foresee. Share market is one of the most impulsive and spot of high premium on the planet. There are no critical techniques exist to foresee the stock cost. Primarily individuals utilize three different ways, for example, major examination, measurable investigation and Machine Learning to foresee the stock cost of offer market yet none of these strategies are demonstrated as a reliably adequate forecast device. Thus, building up a forecast apparatus is one of the difficult undertakings as stock cost relies upon numerous compelling element and highlights. we propose a strong technique to anticipate the offer rate utilizing Moving average based model and contrast how it vary and the genuine cost. For that we gather the share market information of most recent a half year of 5years of various classes, diminish their high dimensionality so it will have the option to prepare quicker and productively and make a similar investigation and our strategy for forecast of following day share cost.To legitimize the adequacy of the framework, diverse test information of companies' stock are utilized to confirm the framework These works show that information mining strategies can be applied for assessment of past stock costs and gain significant data by assessing appropriate monetary the request for the best Moving Average model was discovered to be Further, endeavors were made to figure, as exact as could be normal considering the present situation. Data mining systems can be applied on throughout a wide range of time money related data to make models and further calculations.