Tools Data Mining Techniques and Methods

Tools data mining technique and methods used in developing the proposed study in Sales Forecasting of Starian Marketing Products and the statistical treatment of data in order to complete the study.

Research Methods

There are many methods that can be used in forecasting. In this chapter, the researcher allows finding a suitable model for Starian Marketing Products to forecast the expected future sale of goods. According to Haloub (2013), choosing the appropriate forecasting methods depend on forecasting situation of whether it is a long-term forecasting or short-term forecasting or whether the forecast for a new or existing product. Various univariate time series forecasting models for forecasting seasonal food product sales have been applied in this paper. It compares the out-of-sample forecast accuracy of different models using mean absolute deviation, mean absolute percentage error, and mean percentage error.

The research method chosen focused on finding accountable answers to the research questions. The reviewed literature and studies gave ideas on how to start the forecasting of Starian Marketing products demand and to provide insight into the theoretical/conceptual background of the study. Reviewing related materials helps the researcher to gather valuable data and ideas that can guide in the research. The researchers used the collected data given by the company to conduct the research. The data were the total number of sales from 2012 to 2017.

Research Design

The research design is a systematic investigation of information in order to establish facts and conclusion and to ensure that the evidence obtained enables to effectively address the research problem logically and as unambiguously as possible. According to Green and Tull (1978), A research is the specification of methods and procedures for acquiring the information needed. It is the overall operational pattern or framework of the project that stipulates what information is to be collected from which sources by what procedures.

Experimental Research. Design to be able to predict the phenomenon. It is a systematic approach to research where the researcher manipulates one variable, and control/randomizes the rest of the variables. An experiment is constructed to be able to explain some kind of causation. Experimental research is important to society as it helps us to improve our everyday lives. The researchers try to determine what may occur in the research study of forecasting product sales of Starian Marketing using Time Series Analysis to come up with the results in the ARIMA model.

The researchers used statistical tools to evaluate the criteria of the ARIMA model. The statistical tools are Autoregressive, Difference, Moving average and etc. The researchers used statistical software such as R and Excel to cleanse, detecting outliers from the data and remove inaccurate records. Using statistical tools, the researchers able to test the different factors in separating data to evaluate and forecast the products sales of Starion Marketing.

Historical Data

Historical Data is the collection of data from the past event it includes most data generated either manually or automatically within an enterprise. In business, historical data is important in forecasting because it includes company financial statements, client invoices and any information that has relative predictive value to the future success of the company.

The researchers will gather data from 2012 to 2017 (6 years) to have an accurate source of data. The data includes Product Name, Category, Size, and Sales and etc. which will be cleaned and used in statistical software and tools. The respondents of the study are the owner and employees of Starian Marketing. The respondents’ data from 2012-2017 will be used as training data to discover predictive relationship and test data to assess the strength and acceptability of the model. The researchers will let the data and result be validated by the owner of the company for them to know if it is effective nor can be used for future purposes.

Data Gathering Technique

Data gathering technique is used by the researchers to gather information from the range of sources. It is an important aspect of any type of research study. Inaccurate data collection can impact the results of a study and ultimately lead to invalid results. The researchers used the procedures to gather information for the completion of the study.

Interview. The researchers conducted a structured interview and prepare a standard set of questions to know the answer to the questions needed in gathering the data. Interviews help the researchers uncover rich, deep insight and learn information that they may have missed otherwise.

Documentary Analysis. Documents can provide supplementary research data, making document analysis a useful and beneficial method for most research. It is a technique used to gather requirements during the requirements elicitation phase of a project. It describes the act of reviewing the existing documentation of comparable business processes or systems in order to extract pieces of information that are relevant to the current project and therefore should be considered projects requirements.

Data Mining Technique

Data Mining is about searching large stores of data to uncover patterns and trends that go beyond simple analysis, refers to a process by which patterns are extracted from data. Such patterns often provide insights into relationships that can be used to improve business decision making. Statistical data mining tools and techniques can be roughly grouped according to their use for clustering, classification, association, and prediction.

The researchers used data mining technique forecasting/predicting. The type of forecasting they use is sales forecasting to forecast the sales of Starian Marketing products and to use it for future purposes. Sales forecasting is the process of estimating future sales. Accurate sales forecasts enable companies to make informed business decisions and predict short-term and long-term performance. Companies can base their forecasts on past sales data, industry-wide comparisons, and economic trends. It gives insight into how a company should manage its workforce, cash flow, and resources. In addition to helping a company allocate its internal resources effectively, predictive sales data is important for businesses when looking to acquire investment capital.

To make the research more effective the researchers used time series analysis with the component of seasonal. Time series analysis is a statistical technique that deals with time series data, or trend analysis. Time series data means that data is in a series of particular time periods or intervals. Time series analysis can be useful to see how a given asset, security or economic variable changes over time. It can also be used to examine how the changes associated with the chosen data point compare to shifts in other variables over the same time period.

Statistical Treatment of Data

On this part of the study, the researchers will interpret the gathered data from the company and show it to the chose model which is ARIMA Model in order to come up with an accurate result. Statistical software such as python and excel were deployed for faster and greater reliability of the result. The researchers use the formula

Auto-Regressive Integrated Moving Average (ARIMA). ARIMA model also known as the Box-Jenkins model, is a useful technique used by combining autoregressive (AR) and moving average (MA) process. Given a random process Z, with mean zero and variance. ARIMA methodology attempts to describe the movements in a stationary time series as a function of what is called "autoregressive and moving average" parameters. These are referred to as AR parameters (autoregressive) and MA parameters (moving averages).