BHARATI VIDYAPEETH (DEEMED TO BE UNIVERSITY) COLLEGE OF ENGINEERING DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI,410210 Mini Project Report On Stock Price Prediction Model Subject-: - Machine Learning Presented By Roll No. Name PRN 47 Bhautik Tandel 2143110182 49 Anirudh Jagtap 2143110184 37 Sohail Alekar 2143110185 56 Sakshi Yadav 2143110210 Signature of Internal Examiner BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 1 BHARATI VIDYAPEETH (DEEMED TO BE UNIVERSITY) COLLEGE OF ENGINEERING DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI,410210 This is to certify that the project entitled, “Stock Price Prediction Model”, which is being submitted here with for the award of B.Tech. Department of Computer Science & Business Systems, is the result of the workcompleted by Bhautik Tandel under my supervision and guidance within the four walls of the institute and the same has not been submitted elsewhere for the award of any degree. Guide DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI (Head of Department) DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS KHARGHAR, NAVI MUMBAI Principal DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 2 BHARATI VIDYAPEETH (DEEMED TO BE UNIVERSITY) COLLEGE OF ENGINEERING DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI,410210 This is to certify that the project entitled, “Stock Price Prediction Model”, which is being submitted here with for the award of B.Tech. Department of Computer Science & Business Systems, is the result of the workcompleted by Anirudh Jagtap under my supervision and guidance within the four walls of theinstitute and the same has not been submitted elsewhere for the award of any degree. Guide DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI (Head of Department) DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS KHARGHAR, NAVI MUMBAI Principal DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 3 BHARATI VIDYAPEETH (DEEMED TO BE UNIVERSITY) COLLEGE OF ENGINEERING DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI,410210 This is to certify that the project entitled, “Stock Price Prediction Model”, which is being submitted here with for the award of B.Tech. Department of Computer Science & Business Systems, is the result of the workcompleted by Sohail Alekar under my supervision and guidance within the four walls of the institute and the same has not been submitted elsewhere for the award of any degree. Guide DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI (Head of Department) DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS KHARGHAR, NAVI MUMBAI Principal DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 4 BHARATI VIDYAPEETH (DEEMED TO BE UNIVERSITY) COLLEGE OF ENGINEERING DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI,410210 This is to certify that the project entitled, “Stock Price Prediction Model”, which is being submitted here with for the award of B.Tech. Department of Computer Science & Business Systems, is the result of the workcompleted by Sakshi Yadav under my supervision and guidance within the four walls of theinstitute and the same has not been submitted elsewhere for the award of any degree. Guide DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI (Head of Department) DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS KHARGHAR, NAVI MUMBAI Principal DEPARTMENT OF ENGINEERING & TECHNOLOGY OFFCAMPUS, KHARGHAR, NAVI MUMBAI BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 5 Abstract In this project we will be making a model that will predict the approximate price of a certain stock in the near future. A stock market is a public market where you can buy and sell shares for publicly listed companies. The stocks, also known as equities, represent ownership in the company. The stock exchange is the mediator that allows the buying and selling of shares. Stock Price Analysis is a very nice application of Machine Learning because it involves a lot of previous data and using regression to predict the data through processing it, we will be implementing Machine Learning Algorithms to help the users discover the future value of the stock to gain significant profit. We will be creating a data set for a year’s period and then train the model by applying algorithms until it generates prediction of values that has minimal error. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 6 Index Chapter No. Title Page No. 1 Introduction 8 2 Literature Survey 9 3 System Design 13 4 Implementation 16 5 Results 21 6 Conclusion 22 7 References 23 BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 7 Chapter 1 Introduction Due to the high profit of the stock market, it is one of the most popular investments. People investigated for methods and tools that would increase their gains while minimizing the risk, as the level of trading and investing grew. Two stock exchanges namely- the National Stock Exchange (NSE) and the Bombay Stock Exchange (BSE), which are most of the trading in Indian Stock Market takes place. Sensex and Nifty are the two prominent Indian Market Indexes. Since the prices in the stock market are dynamic, the stock market prediction is complicated. The future price of a stock is the main motivation behind the stock price prediction. In various cases like business and industry, environmental science, finance and economics motivation can be useful. The future value of the company’s stock can be determining. A stock market prediction is described as an action of attempting to classify the future value of the company stock or other financial investment traded on the stock exchange. The forthcoming price of a stock of the successful estimation is called the Yield significant profit. This helps you to invest wisely for making good profits. From gradually the very past years some forecasting models are developed for this kind of purpose and they had been applied to money market prediction. We will be making a mode using Linear Regression. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 8 Chapter 2 Literature Survey Citation Title Authors Methods Advantages Limitations Results [1] The impact of stock market performanc e upon economic growth. Forecasting market price of stock using artificial neural network. Masoud Najeb. Convolut ional Neural Network s Capability of learning on its own. Lot of training data is needed for the CNN to be effective. CNN can effectually recognize the altering trend in stock market price. Mukurte Amod, Tanuja Sarode. Deep Neural Network s More efficient at learning complex features and performing more intensive computationa l tasks It requires very large amount of data in order to perform better than other techniques A cross country empirical analysis on the impact of financial developmen t and asset tangibility on internationa l trade. A survey of attack detection approaches in collaborativ e filtering recommend Hur Jung, Manoj Raj, Yohanes E. Riyanto. LSTMMethod LSTM cells are used in recurrent neural networks that learn to predict the future from sequences of variable lengths It fails to store information for a longer period of time and takes long time to train. These new features are given to a Stacked LSTM Autoencoder for multistepahead estimation of the stock concluding value. The method that is applied on this research to predict Bitcoin on the stock market Rezaimehr, F., & Dadkhah, C. RNN (Recurre nt Neural Network ) RNN enable you to model timedependent and sequential data problems, RNN is tough to train due to the gradient problem [2] [3] [4] BVDU-DET-NM/CSBS/2022-23/ML-miniproject RNN can effectually recognize the altering trend in stock market price. Page | 9 er systems. [5] Research on Sayavong Stock Price Lounnapha. Prediction Method Based on Convolution al Neural Network like stock exchange prediction. Machine Capability of Learning learning on Techniqu its own as the es and human through sentiments is social more media affectively applied. [6] Linear regression Analysis. Hidden Markov Model. [7] Enhancing Soheila Profit by Abrishami. Predicting Stock Prices using Deep Neural Networks Time Series Linear Mode. [8] A Case Study Yahoo Finance Stock Market Ferdiansyah .Wang, D., Liang, Y., Xu, D., Feng, X., & Guan, R. ARIMA Model. [9] Technical analysis and the London stock exchange Chong Terence Tai-Leung, Wing-Kam Ng HoltWinters Seber George AF, Lee Alan J, John Wiley and Sons. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Lot of training data is needed. reflex towards web news is taken into count to reduce the gap and make the prediction much more accurate Hidden Hidden HMM is a Markov Markov stochastic Model gives Model’s main model better problems are assumed to optimization. Evaluation, be a Markov decoding and process with learning. hidden states has more accuracy when compared to other models. It makes the Linear Time series estimation Regression regression is procedure Only Looks a statistical simple. at the Mean method for of the predicting a Dependent future Variable. response based. Only requires Difficult to ARIMA the prior data predict models can of a time turning also be used series to points. to predict the generalize the future price forecast. of your stocks based on the past prices. Capability of Lot of The method learning on training data that is applied its own as the is needed. on human this research sentiments is to predict Page | 10 [10] Classificatio Liaw Andy, n and Matthew regression Wiener by Random Forest. [11] Neural networkbased prediction techniques for global modeling of M Oyeyemi Elijah O., Lee-Anne McKinnell, Allon WV. [12] Forecasting stock market movement direction with support vector machine. Huang Wei, Yoshiteru Nakamori, Shou-Yang Wang [13] Share Price Prediction using Machine Jeevan B. more affectively applied. Random The random Forest forest and Deep algorithm Neural provides a Network higher level Models of accuracy in predicting outcomes over the decision tree algorithm. Method - The multiple deep layers in deep learning neural networks networks allow models to become more efficient at learning complex features and performing more intensive computationa l tasks. Linear Linear regressio regression n fits linearly seperable datasets almost perfectly and is often used to find the nature of the relationship between variables. Artificial ANN learn Neural from events Network and make and decisions BVDU-DET-NM/CSBS/2022-23/ML-miniproject Algorithms are fast to train, but quite slow to create predictions once they are trained. Bitcoin on the stock market Prediction speed is significantly faster than training speed because we can save generated forests for future uses. It requires very large amount of data in order to perform better than other techniques. Deep learning models can achieve stateof-the-art accuracy, sometimes exceeding human-level performance. Data outliers can damage the performance of a machine learning model drastically and can often lead to models with low accuracy. Algorithms are fast to train, but quite slow to While the results produced by linear regression may seem impressive on linearly seperable datasets. neural networks are able to predict Page | 11 Learning Technique Random Forest techniqu es through commenting on similar events. [14] Stock Market Prediction Using Machine Learning Techniques. Naadun Sirimevan. Deep Neural Network s More efficient at learning complex features and performing more intensive computationa l tasks [15] Application of neural network to technical analysis of stock market prediction. Mizuno Hirotaka, Michitaka Kosaka, Hiroshi Yajima, Norihisa Komoda RNN (Recurre nt Neural Network ) RNN enable you to model timedependent and sequential data problems, like stock exchange prediction. BVDU-DET-NM/CSBS/2022-23/ML-miniproject create predictions once they are trained. accurately using the process of backpropagat ion It requires These new very large features are amount of given to a data in order Stacked to perform LSTM better than Autoencoder other for multisteptechniques ahead estimation of the stock concluding value. RNN is tough RNN can to train due to effectually the gradient recognize the problem. altering trend in stock market price. Page | 12 Chapter 3 System Design Stock market prediction is necessary, and it helps investors make the right decision. If a person invested their money in a stock whose future value was decreased, that person would face a loss on his or her investment. Because of that, future stock market predictions are very important. Not only that, but also the human brain cannot make an accurate prediction about the stock market, this is where the need for a technological approach to stock market prediction arises, where a more accurate prediction of the future stock value would result in more profit. Linear regression denotes a linear relationship between the dependent variable and the independent variable or variables. The dependent variable is the target variable in the model. That would be the value that is going to be predicted. In linear regression, “simple linear regression” refers to an analysis that uses one independent variable, whereas “multiple linear regression” refers to an analysis that uses multiple independent variables. Basically, all machine learning algorithms work on historical data where they train the system using those data, and finally, predict the requirements using the trained model. This data set contains a variety of attributes about the stock market. According to the reviews that have been done for stock market prediction using linear regression, there were some common attributes and variables that were contained in the data set related to the stock market as follows: Open Price - Open Price of Stock High Price - Highest price possible at any instance of time Low Price - Lowest Price possible at any instance of time Close Price - Closing Price of Stock Close Price - Closing Price of Stock The accuracy of the result of the model mostly depends on the amount and reliability of the training data set. The below diagram illustrates how the linear regression algorithms are trained for the predictions of the stock market. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 13 Fig. 1 System Architecture This section will critically review the various methodologies that have been used in related research for stock market prediction using linear regression. According to the research¹ conducted by Karim R. and a few others, it ensures that the stock market relies heavily on knowledge whereas, everyone requires a firm grasp of historical and prospective stock market trends nowadays. This could help the investor to get an idea about the stock’s upcoming rates and assess the risk in advance. Furthermore, the main intention of this research is to make a better judgment by using the Linear regression algorithm and Decision tree algorithm with the use of a statistical formula to improve the stock price prediction accuracy. Additionally, this process is required to follow four main steps such as raw data, data cleaning and separating train and test datasets, fitting the train data into the model and evaluating the result to acquire the final outcome and make a decision. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 14 Technology and Language used for project: Languages: 1. Python: Version 3.10.2 2. HTML5, CSS3 Tools with their versions: 1. Operating System: Windows 2. Editors and IDE: a. PyCharm: Version 2022.2.4 b. Visual Studio Code with Jupyter notebook: Version 1.73 3. Dataset: Users, Books and Ratings 4. Libraries and Framework: a. NumPy: Version 1.23.1 b. Pandas: Version 1.4.3 c. Scikit learn: Version 1.1.2 d. Pickle: Version 4 e. Bootstrap: Version 5.2 f. Flask: Version 2.2.2 BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 15 Chapter 4 Implementation Data Summary The dataset is comprised of three csv files: AAPL.csv Stock_dataset: 1. The Required Colums are Open, Low, Close, Volume 2. Date with respect to the dataset Now, we will visualize the Close data values. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 16 BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 17 Linear Regression: 1. Create Training, Testing, Validation Data 2. Apply Linear Regression on Training Data BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 18 3. Evaluate Train Dataset and Apply Linear Regression on Test Data 4. Display Model Results BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 19 BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 20 Chapter 5 Results BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 21 Chapter 6 Conclusion From the research and analysis done so far, we found linear regression is the best approach for the stock prediction problem. By using the past dataset, we have predicted the appropriate right values of stock price. The model we made estimates the price with accuracy of 90% achieved in training as well as testing phase. Libraries such as Pandas, Seaborn, SkLearn, Numpy and Matplotlib are very effective libraries used in making this project execute successfully. The attributes such as open price, close price, day high price, day low price, upper band price, lower band and volume are used to predict the accurate price of the stock. Hence, this model can be used in practical life too to demonstrate the use in actual live markets to earn profits in investment in stocks. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 22 References [1] Masoud Najeb MH, The impact of stock market performance upon economic growth. International Journal of Economics and Financial Issues, 3(4) (2017),pp. 788-798. [2] Murkute Amod,Tanuja Sarode, Forecasting market price of stock using artificial neural network. International Journal of Computer Applications,124 (12) (2015), pp. 11-15 [3] Hur Jung, Manoj Raj, Yohanes E. Riyanto Finance and trade: A cross country empirical analysis on the impact of financial development and asset tangibility on international trade. World Development, 34(10) (2006), pp.1728-1741 [4] Rezaimehr, F., & Dadkhah, C. (2021). A survey of attack detection approaches in collaborative filtering recommender systems. Artificial Intelligence Review, 54(3), 2011-2066. [5] Research on Stock Price Prediction Method Based on Convolutional Neural Network, IEEE 2019- Sayavong Lounnapha. [6] Seber George AF, Lee Alan J. Linear regression Analysis., John Wiley and Sons (2012), p,329 [7] Enhancing Profit by Predicting Stock Prices using Deep Neural Networks, IEEE 2019-Soheila Abrishami. [8] An LSTM-Method for Bit-coin Price Prediction: A Case Study Yahoo Finance Stock Market, IEEE 2019- Ferdiansyah.Wang, D., Liang, Y., Xu, D., Feng, X., & Guan, R. (2018). A contentbased recommender system for computer science publications. Knowledge-Based Systems, 157, 1-9. [9] Chong Terence Tai-Leung, Wing-Kam Ng Technical analysis and the London stock exchange: testing the MACD and RSI rules using the FT30. Applied Economics Letters, 15 (14) (2008), pp. 1111-1114. [10] Liaw Andy, Matthew Wiener Classification and regression by Random Forest. R news, 2 (3) (2002), pp. 18-22. [11] Oyeyemi Elijah O., Lee-Anne McKinnell, Allon WV. Poole Neural network-based prediction techniques for global modeling of M (3000) F2 ionospheric parameter. Advances in Space Research, 39 (5) (2007), pp. 643-650. [12] Huang Wei, Yoshiteru Nakamori, Shou-Yang Wang Forecasting stock market movement direction with support vector machine. Computers & Operations Research, 32 (10) (2005), pp. 2513-2522. BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 23 [13] Share Price Prediction using Machine Learning Technique, IEEE 2019-Jeevan B. [14] Stock Market Prediction Using Machine Learning Techniques, IEEE 2020- Naadun Sirimevan. [15] Mizuno Hirotaka, Michitaka Kosaka, Hiroshi Yajima, Norihisa Komoda Application of neural network to technical analysis of stock market prediction. Studies in Informatic and control, 7 (3) (1998), pp. 111-120 BVDU-DET-NM/CSBS/2022-23/ML-miniproject Page | 24