Browsing by Author "Al Rayhi, Nasser"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Cryptocurrency Exchange Market Prediction and Analysis Using Data Mining and Artificial Intelligence
    (The British University in Dubai (BUiD), 2020-05) Al Rayhi, Nasser
    This thesis discusses the ability of using data mining and artificial intelligence techniques in order to predict crypto currency exchange market prices. The prices of crypto currency are virtual unpredictable and pose numerous liabilities and risks for its investors such as huge sums of money lost in unforeseen variations of crypto currency prices. Furthermore, there are some challenges when it comes to crypto currency prediction such as fake news. Cryptocurrency exchange market have a lot of fake news articles which can affect the market prices by fooling people either to buy or to sell. This research aims to tackle the unpredictability of fluctuating crypto currency prices by researching peer-reviewed articles and determine the applicability of AI algorithm and data mining and its implications for the cryptocurrency market. This thesis will also apply the algorithms in order to determine how applicable it is to predict crypto currency exchange market prices. Furthermore, this thesis discusses blockchain technology in general since this will be the building block for the student research. This thesis is trying to answer the following questions: 1. To which extend can we predict Cryptocurrency exchange market prices using well known prediction algorithms? 2. How accurate is the prediction of the prices? Crypto currency data has been downloaded from different sources and cleansing and normalization methods were used on them. The thesis has examined different algorithms that can be used for prediction such as Moving Average, Auto ARIMA, kNN, Linear Regression and Long Short Term Memory (LSTM). One of the best algorithms in terms of the result is the Long Short Term Memory (LSTM) since it is based on recurrent neural networks which uses loop as a method to learn from heuristics data. For the future improvements can be done to this work by integrating other related data such as social media news using sentimental analysis. Furthermore, more advanced algorithms can be used such as Support Vector machine (SVM) and XGBoost.