Increasing Accuracy of Stock Price Pattern Prediction through Data Augmentation for Deep Learning (데이터 증강을 통한 딥러닝 기반 주가 패턴 예측 정확도 향상 방안)
-
- The Journal of Bigdata
- /
- v.4 no.2
- /
- pp.1-12
- /
- 2019
As Artificial Intelligence (AI) technology develops, it is applied to various fields such as image, voice, and text. AI has shown fine results in certain areas. Researchers have tried to predict the stock market by utilizing artificial intelligence as well. Predicting the stock market is known as one of the difficult problems since the stock market is affected by various factors such as economy and politics. In the field of AI, there are attempts to predict the ups and downs of stock price by studying stock price patterns using various machine learning techniques. This study suggest a way of predicting stock price patterns based on the Convolutional Neural Network(CNN) among machine learning techniques. CNN uses neural networks to classify images by extracting features from images through convolutional layers. Therefore, this study tries to classify candlestick images made by stock data in order to predict patterns. This study has two objectives. The first one referred as Case 1 is to predict the patterns with the images made by the same-day stock price data. The second one referred as Case 2 is to predict the next day stock price patterns with the images produced by the daily stock price data. In Case 1, data augmentation methods - random modification and Gaussian noise - are applied to generate more training data, and the generated images are put into the model to fit. Given that deep learning requires a large amount of data, this study suggests a method of data augmentation for candlestick images. Also, this study compares the accuracies of the images with Gaussian noise and different classification problems. All data in this study is collected through OpenAPI provided by DaiShin Securities. Case 1 has five different labels depending on patterns. The patterns are up with up closing, up with down closing, down with up closing, down with down closing, and staying. The images in Case 1 are created by removing the last candle(-1candle), the last two candles(-2candles), and the last three candles(-3candles) from 60 minutes, 30 minutes, 10 minutes, and 5 minutes candle charts. 60 minutes candle chart means one candle in the image has 60 minutes of information containing an open price, high price, low price, close price. Case 2 has two labels that are up and down. This study for Case 2 has generated for 60 minutes, 30 minutes, 10 minutes, and 5minutes candle charts without removing any candle. Considering the stock data, moving the candles in the images is suggested, instead of existing data augmentation techniques. How much the candles are moved is defined as the modified value. The average difference of closing prices between candles was 0.0029. Therefore, in this study, 0.003, 0.002, 0.001, 0.00025 are used for the modified value. The number of images was doubled after data augmentation. When it comes to Gaussian Noise, the mean value was 0, and the value of variance was 0.01. For both Case 1 and Case 2, the model is based on VGG-Net16 that has 16 layers. As a result, 10 minutes -1candle showed the best accuracy among 60 minutes, 30 minutes, 10 minutes, 5minutes candle charts. Thus, 10 minutes images were utilized for the rest of the experiment in Case 1. The three candles removed from the images were selected for data augmentation and application of Gaussian noise. 10 minutes -3candle resulted in 79.72% accuracy. The accuracy of the images with 0.00025 modified value and 100% changed candles was 79.92%. Applying Gaussian noise helped the accuracy to be 80.98%. According to the outcomes of Case 2, 60minutes candle charts could predict patterns of tomorrow by 82.60%. To sum up, this study is expected to contribute to further studies on the prediction of stock price patterns using images. This research provides a possible method for data augmentation of stock data.
Internet computing is a disruptive IT innovation. Semantic Web can be considered as an IT innovation because the Semantic Web technology possesses the potential to reduce information overload and enable semantic integration, using capabilities such as semantics and machine-processability. How should organizations adopt the Semantic Web? What factors affect the adoption and diffusion of Semantic Web innovation? Most studies on adoption and diffusion of innovation use empirical analysis as a quantitative research methodology in the post-implementation stage. There is criticism that the positivist requiring theoretical rigor can sacrifice relevance to practice. Rapid advances in technology require studies relevant to practice. In particular, it is realistically impossible to conduct quantitative approach for factors affecting adoption of the Semantic Web because the Semantic Web is in its infancy. However, in an early stage of introduction of the Semantic Web, it is necessary to give a model and some guidelines and for adoption and diffusion of the technology innovation to practitioners and researchers. Thus, the purpose of this study is to present a model of adoption and diffusion of the Semantic Web and to offer propositions as guidelines for successful adoption through a qualitative research method including multiple case studies and in-depth interviews. The researcher conducted interviews with 15 people based on face-to face and 2 interviews by telephone and e-mail to collect data to saturate the categories. Nine interviews including 2 telephone interviews were from nine user organizations adopting the technology innovation and the others were from three supply organizations. Semi-structured interviews were used to collect data. The interviews were recorded on digital voice recorder memory and subsequently transcribed verbatim. 196 pages of transcripts were obtained from about 12 hours interviews. Triangulation of evidence was achieved by examining each organization website and various documents, such as brochures and white papers. The researcher read the transcripts several times and underlined core words, phrases, or sentences. Then, data analysis used the procedure of open coding, in which the researcher forms initial categories of information about the phenomenon being studied by segmenting information. QSR NVivo version 8.0 was used to categorize sentences including similar concepts. 47 categories derived from interview data were grouped into 21 categories from which six factors were named. Five factors affecting adoption of the Semantic Web were identified. The first factor is demand pull including requirements for improving search and integration services of the existing systems and for creating new services. Second, environmental conduciveness, reference models, uncertainty, technology maturity, potential business value, government sponsorship programs, promising prospects for technology demand, complexity and trialability affect the adoption of the Semantic Web from the perspective of technology push. Third, absorptive capacity is an important role of the adoption. Fourth, suppler's competence includes communication with and training for users, and absorptive capacity of supply organization. Fifth, over-expectance which results in the gap between user's expectation level and perceived benefits has a negative impact on the adoption of the Semantic Web. Finally, the factor including critical mass of ontology, budget. visible effects is identified as a determinant affecting routinization and infusion. The researcher suggested a model of adoption and diffusion of the Semantic Web, representing relationships between six factors and adoption/diffusion as dependent variables. Six propositions are derived from the adoption/diffusion model to offer some guidelines to practitioners and a research model to further studies. Proposition 1 : Demand pull has an influence on the adoption of the Semantic Web. Proposition 1-1 : The stronger the degree of requirements for improving existing services, the more successfully the Semantic Web is adopted. Proposition 1-2 : The stronger the degree of requirements for new services, the more successfully the Semantic Web is adopted. Proposition 2 : Technology push has an influence on the adoption of the Semantic Web. Proposition 2-1 : From the perceptive of user organizations, the technology push forces such as environmental conduciveness, reference models, potential business value, and government sponsorship programs have a positive impact on the adoption of the Semantic Web while uncertainty and lower technology maturity have a negative impact on its adoption. Proposition 2-2 : From the perceptive of suppliers, the technology push forces such as environmental conduciveness, reference models, potential business value, government sponsorship programs, and promising prospects for technology demand have a positive impact on the adoption of the Semantic Web while uncertainty, lower technology maturity, complexity and lower trialability have a negative impact on its adoption. Proposition 3 : The absorptive capacities such as organizational formal support systems, officer's or manager's competency analyzing technology characteristics, their passion or willingness, and top management support are positively associated with successful adoption of the Semantic Web innovation from the perceptive of user organizations. Proposition 4 : Supplier's competence has a positive impact on the absorptive capacities of user organizations and technology push forces. Proposition 5 : The greater the gap of expectation between users and suppliers, the later the Semantic Web is adopted. Proposition 6 : The post-adoption activities such as budget allocation, reaching critical mass, and sharing ontology to offer sustainable services are positively associated with successful routinization and infusion of the Semantic Web innovation from the perceptive of user organizations.
Recently, AlphaGo which is Bakuk (Go) artificial intelligence program by Google DeepMind, had a huge victory against Lee Sedol. Many people thought that machines would not be able to win a man in Go games because the number of paths to make a one move is more than the number of atoms in the universe unlike chess, but the result was the opposite to what people predicted. After the match, artificial intelligence technology was focused as a core technology of the fourth industrial revolution and attracted attentions from various application domains. Especially, deep learning technique have been attracted as a core artificial intelligence technology used in the AlphaGo algorithm. The deep learning technique is already being applied to many problems. Especially, it shows good performance in image recognition field. In addition, it shows good performance in high dimensional data area such as voice, image and natural language, which was difficult to get good performance using existing machine learning techniques. However, in contrast, it is difficult to find deep leaning researches on traditional business data and structured data analysis. In this study, we tried to find out whether the deep learning techniques have been studied so far can be used not only for the recognition of high dimensional data but also for the binary classification problem of traditional business data analysis such as customer churn analysis, marketing response prediction, and default prediction. And we compare the performance of the deep learning techniques with that of traditional artificial neural network models. The experimental data in the paper is the telemarketing response data of a bank in Portugal. It has input variables such as age, occupation, loan status, and the number of previous telemarketing and has a binary target variable that records whether the customer intends to open an account or not. In this study, to evaluate the possibility of utilization of deep learning algorithms and techniques in binary classification problem, we compared the performance of various models using CNN, LSTM algorithm and dropout, which are widely used algorithms and techniques in deep learning, with that of MLP models which is a traditional artificial neural network model. However, since all the network design alternatives can not be tested due to the nature of the artificial neural network, the experiment was conducted based on restricted settings on the number of hidden layers, the number of neurons in the hidden layer, the number of output data (filters), and the application conditions of the dropout technique. The F1 Score was used to evaluate the performance of models to show how well the models work to classify the interesting class instead of the overall accuracy. The detail methods for applying each deep learning technique in the experiment is as follows. The CNN algorithm is a method that reads adjacent values from a specific value and recognizes the features, but it does not matter how close the distance of each business data field is because each field is usually independent. In this experiment, we set the filter size of the CNN algorithm as the number of fields to learn the whole characteristics of the data at once, and added a hidden layer to make decision based on the additional features. For the model having two LSTM layers, the input direction of the second layer is put in reversed position with first layer in order to reduce the influence from the position of each field. In the case of the dropout technique, we set the neurons to disappear with a probability of 0.5 for each hidden layer. The experimental results show that the predicted model with the highest F1 score was the CNN model using the dropout technique, and the next best model was the MLP model with two hidden layers using the dropout technique. In this study, we were able to get some findings as the experiment had proceeded. First, models using dropout techniques have a slightly more conservative prediction than those without dropout techniques, and it generally shows better performance in classification. Second, CNN models show better classification performance than MLP models. This is interesting because it has shown good performance in binary classification problems which it rarely have been applied to, as well as in the fields where it's effectiveness has been proven. Third, the LSTM algorithm seems to be unsuitable for binary classification problems because the training time is too long compared to the performance improvement. From these results, we can confirm that some of the deep learning algorithms can be applied to solve business binary classification problems.