Generally, feature selection techniques designed with different appraisement criteria, are categorized into two divisions: the filter algorithms (Huang and Tesai, 2009;Kara et al ,2011), and the wrapper algorithms(Liang et al,2011;Nair et al,2010). Gaining no feedback from classifier, the filter algorithms estimate the classification performance by some indirect assessments, such as distance measures that reflect how well the classes separate from each other (Li-Ping et al, 2011).On the contrary, the wrapper algorithms are classifier-dependent. Based on the classification accuracy, the methods evaluate the ‘‘goodness” of the selected feature subset directly that should intuitively yield better performance. Many experimental results reported so far, are in favor of the wrapper methods (Luo and Chen, 2013;Teixeria and Olivieria,2010). In spite of the good performance, the wrapper methods have had restricted employment because of the high computational intricacy involved. In this paper genetic algorithm (GA) as a filter technique is employed in terms of feature selection purpose to result in a better diagnosis of the stock’s trends.
3.3. Support vector machine (SVM)
Support vector machine (SVM), which is based on structural risk minimizations concepts and statistical learning theory, was developed firstly by Vapnik (1995).Two remarkable applications of SVM is pattern recognition (classification) as well as regression estimation (approximation of functions).These applications make SVM as a popular method among researchers to implement it to solve problems such as nonlinear modeling, time series forecasting and etc. In this article, it’s classification algorithm has been applied to forecast the price of each stock in certain time peri...
... middle of paper ...
...diction. When using the RBF kernel, there are two notable parameters C, γis needed to be carefully chose in the SVM algorithm.In this article, the parameters of C, γ are optimized by cross-validation (Lin et al, 2008) and grid search. The training set is divided into six folds. One fold was taken as the validation set and the others applied for training. The grid point with the highest precision of predicting is used as the value of the two parameters. Rapid Miner software SVM’’s operator was used for conduction of SVM algorithm. The original data (selected technical indicators resulted by GA) are scaled into the range of (0, 1). The purpose of linear scaling is to autonomously normalize each feature quantity to the specified range. It guarantees that the larger value input attributes do not overwhelm smaller value inputs; hence helps to decrease prediction errors.
The first financial ratio of the analysis is the Price to Earnings ratio (“P/E ratio”). The ratio is computed by dividing the price of one share of common stock, by the earnings per share of common stock. This analysis uses diluted earnings per share which assumes the issuance of new stock for all existing stock options. Also, the price of the stock was computed as an average of the fourth quarter high and low stock prices published in the 10K report of each company, because the year end stock prices were not listed for all the companies. Because the P/E ratio measures the relative costliness of different stocks, in relation to their income, it provides a useful place to begin the analysis.
The estimates of cost of capital for equity 6.14% are making by using the capital asset pricing model (CAPM) to generate forecast of DDM and RIM. This method is defined by the sum of risk free rate plus beta that multiplied with a risk premium. Particularly, the beta, which is a quantitative measure of the volatility of company stock relative to the unstable of the overall market, found in JB HI-FI case at 0.56 (JB HI-FI financial statement 2016). It
The second model is the LEARN model which was developed by Dr. Elois Ann Berlin and Dr. William Fowkes (1983). It has similar objectives to the previous pneuomnic, which are to provide health care practitioners with a simple pneumonic to improve cross-culture communication. This model should not be viewed as a separate pneumonic to memorize but rather a complimentary method to implement culturally appropriate health care for all patients.
Classification Text documents are arranged into groups of pre-labeled class. Learning schemes learn through training text documents and efficiency of these system is tested by using test text documents. Common algorithms include decision tree learning, naive Bayesian classification, nearest neighbor and neural network. This is called supervised learning.
... (such as linear regression and logistic regression) or forecasting techniques (such as neural networks and survival analysis) can be used to determine a given result, such as demand forecasting or customer attraction through direct marketing and evaluate their responses.
The MDA model also showed potential to ease some problems in the selection of securities for a portfolio, but further investigation was recommended.
The project is done to find out the impact of stock split on the stock market. In our project, we have made use of event study methodology to assess the accuracy of stock price reaction of 39 public listed Indian companies in National Stock Exchange (BSE) in the year 2006 and onwards. The abnormal returns (actual returns-returns from regression line) results were taken for 20 days before and after the announcement date to test whether the result is significant or not (Level of significance=5%). The project shows that there is no significance difference in the price level before the announcement date while after the announcement date, there was a significant difference in the price level for few days(level of significance being 5%) The project supports the hypothesis that Indian stock market is semi strong efficient.
Learning is defined as a, "relatively permanent change in an organism's behavior due to experience" (Myers, 2010). The purpose of this paper is to present a critical analysis of the different theoretical approaches and explanations for learning through an examination of the theories of behaviorism, social learning and cognitive. I will investigate the principles and postulates of each theory, their strengths and their weaknesses. It is my belief that because each theory is best applicable to varying types of learning, it is best that a combination of each is used to provide the most complete learning experience.
Following the trend of economy, it is important to investors to understand that strong economy creates strong stock market. To elaborate further, as stock prices are increased by current and future expectations of earnings, thus without a strong economy it would be difficult for the companies to increase and sustain their earnings (Kong 2013). The economy development is usually calculated using the gross domestic product of a countries. On the other hand, a change is the stock price can also cause a major impact to the consumers and investors directly. Hence, a loss in confidence by investors can cause a downturn in consumer spending in the long term, which will also affect the economy’s output (Aysen 2011). The graph below shows the relationship of stock market price (KLCI) and the GDP of Malaysia in 2009. Thus, it can be concluded that the economy and the stock market has a positive relationship.
... applied on different Domain data sets and sub level data sets. The data sets are applied on Maximum entropy, Support Vector Machine Method, Multinomial naïve bayes algorithms, I got 60-70% of accuracy. The above is also applied for the Unigrams of Maximum entropy, Support Vector Machine Method, Multinomial naïve bayes algorithms achieved an accuracy of 65-75%. Applied the same data on proposed lexicon Based Semantic Orientation Analysis Algorithm, we received better accuracy of 85%. In subjective Feature Relation Networks Chi-square model using n-grams, POS tagging by applying linguistic rules performed with highest accuracy of 80% to 93% significantly better than traditional naïve bayes with unigram model. The after applying proposed model on different sets the results are validated with test data and proved our methods are more accurate than the other methods.
Stock market prediction is the method of predicting the price of a company’s stock. It is believed that stock price is lead by random walk hypothesis. Random walk hypothesis states that stock market price matures randomly and hence can’t be predicted. Pesaran (2003) states that it is often argued that if stock markets are efficient then it should not be possible to predict stock returns. In fact, it is easily seen that stock market returns will be non-predictable only if market efficiency is combined with risk neutrality. On the other hand it is also been concluded that using variance ratio tests long horizon stock market returns can be predicted....
Business forecasting can be used in a wide variety of contexts, and by a wide variety of businesses. For example, effective forecasting can determine sales based on attendance at a trade show, or the customer demand for products and services (Business and Economic Forecasting, p.1). One of the most important assumptions of business forecasters is that the past acts as an important guide for the future. It is important to note that forecasters must consider a number of new information, including rapidly changing economic conditions and globalization, when creating business forecasts based on past sales.
I am currently majoring in Finance Management. Most of the time people think of finance as just managing money. However, finance is needed for so much more! The finance industry deals with starting businesses, developing new products, expanding markets, as well as everyday things like saving for retirement, purchasing a home, and even insurance. The stock market, asset allocation, portfolio analysis, and electronic commerce are all key aspects in finance. In this paper, I will explain how these features play a vital role in the industry, along with the issues that come with these factors.
Humans can expand their knowledge to adapt the changing environment. To do that they must “learn”. Learning can be simply defined as the acquisition of knowledge or skills through study, experience, or being taught. Although learning is an easy task for most of the people, to acquire new knowledge or skills from data is too hard and complicated for machines. Moreover, the intelligence level of a machine is directly relevant to its learning capability. The study of machine learning tries to deal with this complicated task. In other words, machine learning is the branch of artificial intelligence that tries to find an answer to this question: how to make computer learn?
In the modern world, financial markets play a significant role, with huge volumes of everyday dealings. They form part of contemporary economic lifestyle and determine the level of success of many people. Humans have always been uncertain of what the future holds and thus, tried to forecast it. The forecast of course cannot omit the likelihood of “easy money” by forecasting the prices of equity markets in the future.