ANALYSIS OF STOCKS MOVEMENTS USING DEEP LEARNING

• Main Author: Harapan Pangihutan, Rio
• Format: Final Project
• Language: Indonesia
• Subjects: Fisika
• Online Access: https://digilib.itb.ac.id/gdl/view/41781
• Institution: Institut Teknologi Bandung

Econophysics is one branch of interdisciplinary science in physics that applies theories and methods developed by physicists to solve problems in the eco- nomic field. One of the physics approaches used in economic problems such as the use of statistical physics to conduct studies on the capital market by studying stock market movements by adopting statistical physics and com- putational physics approaches. Shares are units of value or bookkeeping in various financial instruments that refer to the ownership of a company. The stock price movement will be very dynamic every second. To predict this can use two methods of analysis, the first is fundamental analysis that focuses on the analysis of financial statements and company conditions, then the second is technical analysis carried out by reading the movement of stock prices by looking at the graph and calculation of historical stock price data. Technical analysis can be done using Deep Learning. Deep Learning or Representation learning learns how to represent data so that it can be easily extracted to re- trieve important information in making a prediction or classification machine. Using Deep Learning will produce a machine predicting stock price movements with certain accuracy and accuracy.