WRITING CENTER GUIDE
DATA ANALYTICS TO FORECAST THE MARKET STOCK PRICE
INTRODUCTION
Data Analytics is a tract which studies some raw data obtained from certain sources and results the conclusions of such data [1]. The conclusion can be such as trend or prediction in order to optimize or increase the efficiency of business or system [1]. Besides that, Data Analytics is also used to describe the analysis of large volumes of data [2]. It considers historical data in order to find the better solutions for the complex business problems [3]. However, there is another term that most of people sometimes misunderstand, called by Big Data, which is a huge size of raw data and still unstructured, that comes from diverse sources [3]. Data Analytics has a specific goal in mind to look for ways using some techniques to gain support, while big data is a collection of a great deal of data which need plentiful filter to obtain beneficial knowledge from it [3]. So far, Data Analytics has already implemented in various fields such as in health care, disaster, child welfare, physical security, and others [4]. However, this brochure encompasses the implementation of Data Analytics for doing market stock prediction.
FUNDAMENTAL ASSUMPTION
There are four basic types of Data Analytics [1]. The first one is Descriptive Analytics [1]. This first type which describes the things that already happened in a period of time [1]. Also, it creates a summary of historical data in order to yield some possibly and useful information for the further analysis [5]. The second one is Diagnostic Analytics [1]. It focuses more on why something happened, and involves more diverse data inputs and some hypothesis [1]. Sometimes it answers questions such as “Did the weather affect beer sales?” or “Did the latest marketing campaign has impacts to the sales?” [1]. The third one is Predictive Analytics [1]. Such as the name, which is predictive, it tells what is probably will happen in the near term [1]. In other words, predictive analytics is the utilization of data, statistical algorithms, and machine learning techniques to identify the possibility of future result based on the given historical data [6]. The fourth one is Prescriptive Analytics [1]. The goal of this basic type is to optimize a set of decisions for directing a given business situation [7]. Also, it answers the question such as “what should we do?” after doing the first three basic types of Data Analytics [7].
Implementing Data Analytics field in the world of stock market might give much benefits especially for the trader since probably it can reduce the lost and increase the profit. The definition of stock market is the collection of market and exchange where activities such as buying, selling, and issuance of shares take place [8]. In this case, since the utilization of the Data Analytics is for predicting the movement of the market stock price, the trader who does the activities of buying, selling, and issuance should obtain the information from the data scientist in order to minimize the lost and maximize the profit of doing the market stock activities. A data scientist is an individual who performs statistical analysis, data mining, and retrieval process on large amount of data to identify trends, figures, and other relevant information [2].
RESEARCH QUESTION
While doing research in Data Analytics, there are questions that might emerge. The questions might come from the researchers, or maybe can be asked by the readers who want to implement the method of the research.
Since the research is specifically for the implementation of Data Analytics in order to predict the movement price of market stock, the research questions might be such as the following:
- Does the method can be used in the index market? Also, does the method is applicable for all companies share?
Such question usually emerges to make the research more specific. Whether the method only can be implemented in one country or more? Also, whether the method is only for the blue chips, or can be used to predict the price of the penny stock?
- How many parameters does the method needs in order to gain the desired result?
Parameters can be such as the name of the share, or maybe the initial price, or maybe the duration of time of holding the shares.
- How accurate the result is?
Since prediction will not gain the exact value, the information of the percentage of accuracy is required in order to measure or to know the precise level.
- What or how many methods does it take?
Every research should have methods. However, the number of methods used during the research might be different to others. Also, the type of method used for the research should be different as well.
- Is there any proof whether the method had successfully met the expectation result?
This question regularly can be answer to the result of the research. Indeed, the purpose of this question is to know whether the search succeed or vice versa.
CREDIBLE EVIDENCE
Credible evidence of the research in implementing Data Analytics for predicting the market stock price movement are the data source and the techniques. Both will provide better evidence for decision making [9].
A good decision making will gain the relevant desired goals [9]. The result should possible to address the problem. Therefore, the result automatically can also be one of the credible evidence parts since the results emerge after the source of data is proceeded by the techniques which are used to predict the final value.
RESEARCH METHODOLOGIES
Since this brochure discusses specifically in implementing Data Analytics in order to predict the stock market price, the methodologies used to do so will be more specific as well. Therefore, the methodologies, which relate to the Data Analytics field, encompass all Technics such as the following [10]:
- Statistical
There are three types for doing prediction using statistical methodology [10]. The three types are ARIMA, ESN, and Regression [10]. The ARIMA is good for time series data in order to forecast the future point [10]. The rest of the types, such as ESN and Regression, depict another group of statistic approach which utilizes multiple input of variables [10].
- Pattern Recognition
This type of model focus on the detection of patterns and trends [10]. It involves the visual analysis to show variation in price, volume, or other indicators such as price momentum [10]. The pattern has the capacity to inform investors about the future stock evolution [10].
- Machine Learning
This model is classified into supervised and unsupervised learning [10]. The supervised learning goals’ is to train an algorithm to automatically match the input data to the output data [10]. Meanwhile, the unsupervised learning goals’ is to train algorithms to find a pattern, relationship, or group in the given data set [10].
- Sentiment Analysis
This model works by analyzing the text such as news or tweets of the specific stock and public companies [10]. It can improve the efficiency of models to predict volatility trends in the stock market [10].
- Hybrid
This is a combination model [10]. It harnesses more than one model such as machine learning and sentiment analysis to predict the stock movement [10]. Also, it can combine the two techniques such as Statistical and Pattern Recognition as well to do the prediction
SOURCES OF DATA ANALYTICS IN STOCK MARKET PREDICTION
The sources to do the research about Data Analytics which specifically does the research in predicting the market stock movement can be classified into three part such as the following:
Scholarly Journals
The scholarly journals in Data Analytics for Market Stock Prediction are the writings of Data Analytics that use methods in order to do some prediction of Market Stock Price Movement, and also use the reliable source of information [11]. The writings are done by the professionals or someone who is sophisticated in such field [11]. The content should comprise the Data Analytics techniques and the sample of name of the share and the price information as the object of the research. The scholarly journal for this field ca be shown as the following list:
- International Journal of Data Science and Analytics. (2016-2019). Switzerland: Springer International Publishing.
- The Journal of Finance and Data Science. (2015-2019). Southwest Jiaotong University, China & Oklahoma State University, US: China Science Publishing & Media Ltd.
- Expert Systems with Application: An International Journal. (2019). Elmsford, NY: Elsevier Ltd.
- International Journal of Computer Science and Information Security. (2011- 2019). Pittsburgh, PA: IJCSIS Publication.
- Academy of Information & Management Sciences Journal. (2000). United States: Allied Academies.
Reference Guide
Prior to doing the research, learning several components through the reference guide be helpful. It will help researchers or users to understand how to use the specific report’s data [12]. Here are the several Reference Guides that might help the researchers prior to doing the research:
- Shah, H. Isah and F. Zulkernine, “Stock Market Analysis: A Review and Taxonomy of Prediction Techniques,” International Journal of Financial Studies, 2019.
- Power, R. Roth and D. Cyphert, “Analytics and Evidence-Based Decision Making,” in AMCIS 2018 Proceedings, 2018.
- Sivarajah, M. M. Kamal, Z. Irani and V. Weerakkody, “Critical analysis of Big Data challenges and analytical methods,” Journal of Business Research, vol. 70, p. 263–286, 2017.
- Sismanoglu, F. Kocer, M. A. Onde and O. K. Sahingoz, “Deep Learning Based Forecasting in Stock Market with Big Data Analytics,” in IEEE, Istanbul, Turkey, Turkey, 2019.
- Wen, P. Li, L. Zhang and Y. Chen, “Stock Market Trend Prediction Using High-Order Information of Time Series,” in IEEE Access, vol. 7, pp. 28299-28308, 2019.
doi: 10.1109/ACCESS.2019.2901842
Databases
The source of journal or article that would assist researchers to do the research in Data Analytics field can be found at following database:
- IEEE Xplore
- ASME Conference Proceedings (Current)
- Access Engineering
The source of information about historical market stock data can be found through the following database:
- Global Finance Data
- Tick Data
- Bloomberg
- Reuters
Popular Magazines
The popular reading or magazine of Data Analytics and the updates of new technology to analyze or do prediction can be shown by the following [13]:
- Analytics insight
- Emerj
- Datafloq
- Dataconomy
- Dataversity
WRITING STRUCTURES & GENRES
The general writing structure for Data Analytics report can be such as the following [14]:
Introduction
This part summarizes the study and data. Also, it usually contains the background of the study. In the other words, it briefly outlines the answer of the big questions why the paper is written [14].
Body
This part can be organized in two ways such as the following [14]:
Traditional
This way is divided into several section such as the following [14]:
- Data
- Methods
- Analysis
- Results
This format is a general form in the research papers. In a data analysis paper, you should describe the analyses that you performed [14].
Question-Oriented
In this format there is a single Body section, usually called “Analysis”, and also there is a subsection for each question raised in the introduction, usually taken in the same order as in the introduction [14].
Conclusion/Discussion
The conclusion should comprise the questions and conclusions of the introduction, perhaps augmented by some additional observations or details gleaned from the analysis section [14]. New questions, future work, etc., can also be raised here [14].
Appendix/Appendices
One or more appendices are the place to out details and additional materials [14]. These might include such items as the following [14]:
- Technical descriptions of (unusual) statistical procedures
- Detailed tables or computer output
- Figures that were not central to the arguments presented in the body of the report
- Computer code used to obtain results.
In addition, the above structure can be applied in all Data Analytics sub-field. The writing of Data Analytics in forecasting market stock prices, indeed, can use the above structure since that is the general structure of engineering writings.
In Data Analytics writing, there are several genres that might fit to it. The writing genres can be such as the following:
Article
As shown through scholarly journal, this type of genre can be applied in order to write about the new implementations or practice of Data Analytics, indeed, can be applied as well for the Data Analytics in market share forecasting.
Essay
In the college world, especially for Data Analytics students, this genre is generally useful in order to answer or fulfill the given assignments.
Thesis
Nearly similar to the essay, however this genre is more complete in terms of methodology. It usually contains more rigid part such as the methodology testing and implementations
Procedural
This genre can also be applied for Data Analytics in forecasting the market stock. It might contain the step by step manual in order to show how the prediction techniques is used. This writing genre can help the beginners to apply the techniques prior to doing their research.
Lab Report
This genre is written to describe and analyze a laboratory experiment [15]. In Data Analytics field, especially to forecast the market stock price movement, this type of genre can be applied since it helps the researchers to write all the data collected, such as the text part or even the equation, during the experiment.
CURRENT PRESSING ISSUES
The Pressing Issues in Data Analytics are diverse. However, since this handout focuses on the market stock implementation, the following current pressing issues are the critical points of data analytics that relates more to the market stock analysis field.
- Stock Market Data API
Stock market data APIs offer real-time or historical data on financial assets that are currently being traded in the markets [16]. These data can be used for generating technical indicators which are the foundation to build trading strategies and monitor the market [16]. However, the numbers of API platforms are many and keep increasing. Therefore, the data scientist and IT Developer should always be up to date in order in terms of API platform in order to enhance or develop the system relates to the Stock Market Data API.
- Lack of Data Scientist Experts
Since the number of data in this world is increasing, the number of problems that need to be solved is increasing as well. However, the number of people who are expert in analyzing the data in order to do the forecasting is limited as well [17]. Therefore, the number of experts in Data Analytic which is not as much as the market demands might be part of the pressing issues.
- Data Protection
Selecting only the necessary data in to do analytics is important in order to comply the regulation [19]. Whenever possible, remove personally identifiable or sensitive information before adding it to your databases [19]. Otherwise, personally identifiable or sensitive information can be misused by the irresponsible party.
- The Limitation of Database Engine Capacity
Since usually running the big data requires excellent databases, choosing a proper database engine might result additional cost. In this case, database engine should able to deliver data in a good speed [20]. Also, it should offer comprehensive solutions for any robust data processing solution at large-scale database management environment [20].
- Data Quality
In order to gain the better result of forecasting, having good quality of data is a must. Sometimes, more cost might be spent in order to have a good quality of data. Otherwise, the result will not approach the maximum level of precise, since Data Quality-related problems always surfaced in “historical data,” which may have been gathered from multiple sources with inconsistent standards and varying levels of accuracy [18].
References
[1] | Investopedia, “Corporate Finance & Accounting,” Investopedia, 27 April 2019. [Online]. Available: https://www.investopedia.com/terms/d/data-analytics.asp. [Accessed 2019 November 13]. |
[2] | Informatica, “What is Data Analytics?,” Informatica, 2019. [Online]. Available: https://www.informatica.com/services-and-training/glossary-of-terms/data-analytics-definition.html#fbid=cu5tC8MSB-X. [Accessed 14 November 2019]. |
[3] | TALENTEDGE, “Difference Between Big Data and Data Analytics,” TALENTEDGE, 2019. [Online]. Available: https://talentedge.com/blog/difference-between-big-data-and-data-analytics/. [Accessed 14 November 2019]. |
[4] | Capella University, “Four Example of Data Analytics in Action,” Capella University, 2019. [Online]. Available: https://www.capella.edu/rio/it-hub/analytics-careers/big-data-and-analytics/. [Accessed 14 November 2019]. |
[5] | WhatIs.com, “Descriptive Analytics,” TechTarget, 2019. [Online]. Available: https://whatis.techtarget.com/definition/descriptive-analytics. [Accessed 14 November 2019]. |
[6] | SAS, “Predictive Analytics,” SAS Institute Inc, 2019. [Online]. Available: https://www.sas.com/en_us/insights/analytics/predictive-analytics.html. [Accessed 14 November 2019]. |
[7] | Gurobi Optimization, “The Power of Analytics,” Gurobi Optimization, [Online]. Available: https://www.gurobi.com/company/about-gurobi/prescriptive-analytics/. [Accessed 14 November 2019]. |
[8] | Investopedia, “Stock Market,” Investopedia, 25 June 2019. [Online]. Available: https://www.investopedia.com/terms/s/stockmarket.asp. [Accessed 14 November 2019]. |
[9] | D. Power, R. Roth and D. Cyphert, “Analytics and Evidence-Based Decision Making,” in AMCIS 2018 Proceedings, 2018. |
[10] | D. Shah, H. Isah and F. Zulkernine, “Stock Market Analysis: A Review and Taxonomy of Prediction Techniques,” International Journal of Financial Studies, 2019. |
[11] | Cornel Univeristy Library, “Distinguishing Scholarly from Non-Scholarly Periodicals: A Checklist of Criteria: Introduction and Definitions,” Cornel Univeristy, 31 May 2019. [Online]. Available: http://guides.library.cornell.edu/scholarlyjournals. [Accessed 16 November 2019]. |
[12] | IGI Global, “IGI Global – Disseminator of Knowledge,” IGI Global, 1988 – 2019. [Online]. Available: https://www.igi-global.com/dictionary/data-system-embedded-guidance-significantly-improves-data-analyses/50509. [Accessed 16 November 2019]. |
[13] | IndustryWired, “IndustryWired,” Stravium Intelligence LLP, 2018. [Online]. Available: https://industrywired.com/top-10-big-data-and-artificial-intelligence-magazines-and-publications/. [Accessed 16 November 2019]. |
[14] | B. Junker, “Carnegie Mellon University – Statistic & Data Science,” [Online]. Available: http://www.stat.cmu.edu/~brian/701/notes/paper-structure.pdf. [Accessed 16 November 2019]. |
[15] | Massey University, “What is a lab report?,” Massey University, 1998-2010. [Online]. Available: http://owll.massey.ac.nz/assignment-types/what-is-a-lab-report.php. [Accessed 16 November 2019]. |
[16] | S. Huang, “Best 5 free stock market APIs in 2019,” Towards Data Science, 14 November 2019. [Online]. Available: https://towardsdatascience.com/best-5-free-stock-market-apis-in-2019-ad91dddec984. [Accessed 16 November 2019]. |
[17] | insideBIGDATA, “Infographic: The Data Scientist Shortage,” insideBIGDATA, 19 August 2018. [Online]. Available: https://insidebigdata.com/2018/08/19/infographic-data-scientist-shortage/. [Accessed 17 November 2018]. |
[18] | P. Guha, “Challenges of Data Quality in the AI Ecosystem,” DATAVERSITY, 12 November 2019. [Online]. Available: https://www.dataversity.net/challenges-of-data-quality-in-the-ai-ecosystem/. [Accessed 17 November 2019]. |
[19] | bigdata, “Securing Your Data with AWS,” big data analytics news, 7 September 2019. [Online]. Available: https://bigdataanalyticsnews.com/securing-your-data-with-aws/. [Accessed 17 November 2019]. |
[20] | bigdata, “The Limitation of MySQL Database in a Typical Big Data Environment,” big data analytics news, 10 May 2019. [Online]. Available: https://bigdataanalyticsnews.com/limitation-of-mysql-database-in-big-data/. [Accessed 17 November 2019]. |