Expected lifetime of the product
Software evolution process does not end at the death of a software system but usually continues its evolution over generations through being replaced by newly built software.
Software lifetime is depending on the usage of company for example- In present, company is using the software for only stock market data prediction but after some time in future it also wants to use that software for data prediction as well as for maintaining the company’s data and record of every employee so in that case changes in the software will be required or they must change the software. So basically, life time of a software depends on the needs of the company.
Our program is based on an algorithm and it produce the output based on that algorithm. There are number of factors that might affect the lifetime of our product.
If user wants to predict the data using different algorithm.
If there is another prediction program with better prediction rates.
If the program does not fit in hardware environment.
Our program work with Python language and if user is not comfortable with python then they might switch the program.
Reliability & Performance requirements
Product reliability is defined as the probability that a device will perform its required function for a specific period. Our client is a Trading company and they are going to take the decisions based on the prediction of our product and if the product is not reliable then company may not be able to take the right decision. There are some features of a reliable product.
Product should be easy to install and operate.
It should give the accurate prediction.
Product should not crash during the prediction of a data.
The output should be understandable. (Everyone can easily understand the prediction rate)
Product should easily adopt the exiting hardware environment of a company.
Performance is an effective trait of a product. In our case, performance of our product is directly connected to the decision-making process of a company because they are going to take the decision based on the prediction of our product.
Product should be able to give accurate prediction on a daily basis.
It should be able to run heavy and complex data file.
It should take appropriate time to run a data file.
It should be able run different type of data file like: Live data, Structure data, Unstructured data etc.
Existing software functions, data, and hardware environments
Companies are using these software’s currently for stock market prediction:
Gephi - Gephi is also an open-source network analysis and visualization software package written in Java on the Net-beans platform.
Data-wrapper - Data-wrapper is an online data-visualization tool for making interactive charts. Once you upload the data from CSV/PDF/Excel file or paste it directly into the field, Data-wrapper will generate a bar, line, map or any other related visualization.
Solver - Solver specializes in providing world-class financial reporting, budgeting and analysis with push-button access to all data sources that drive company-wide profitability.
Qlik - Qlik lets you create visualizations, dashboards, and apps that answer your company’s most important questions. Now you can see the whole story that lives within your data.
Data
Company has employee data, stock market data, live data, structured data, unstructured data, and historical data in their database system. Company needs to store this data to run company successfully.
Hardware
Company has desktop computers and provided laptop to employees. All the devices relate to main hub through star topology. Company is using cloud storage to store huge data. Two high performance servers are used by companies with 256 GB RAM.
Future extensions of the product
To get the best outcomes or output from the product we can improve it by adding some new features as well as by adding new algorithms which can predict the data very efficiently.
We know that the data is not remain the same for any company and by time some new parameters or problems come into picture. In that case updating of the product, algorithm and system is required.
Sometime company’s update their hardware or system and in this case, product should be easily able to adopt this change by adding some new features in the product.
Required implementation environment
Below are the steps that we need to implement to achieve the best output from our program.
Web service to collect the data.
Python or R programming language.
PyCharm (IDE) Software for data analysis.
Data sorting to make data clean.
Different algorithms for best outcomes.
Window operating system.
Feasibility of proposed solutions (method)
The methodology of our project is simple. We have to predict the stock market using different algorithm. We are going to collect the real-time data of a company from “Kaggle.com” and analyze that data using different algorithms. We will use python to clean our data set. Working with python language “PyCharm” is the appropriate software because it installs all the required libraries automatically. Artificial neural networks (ANN), Support vector and KNN are some of the best fit algorithms for prediction. Implement all of them on data set to find the best fit model.
Training, installation, and documentation requirements
Training: After implementation of our stock market prediction program, it is necessary to train the users who will use this program for prediction. Proper training can make workers more efficient. If employees can’t effectively use the newly implemented solutions, productivity and employee morale will suffer. For our program training will be easy because we are designing this program in python and mostly IT departments are familiar with python environment.
In their training user are going to learn each step of the program like: How to collect the data, how to implement the program, what king of data is more productive? And how to get optimum output from a data set. Training is an important step to get the best result from product.
Installation
This program will be installed in company’s machines in which their data is stored from different sources for prediction of stock market. The installation will take some time; it depends upon the system requirements and configuration. This program will be python based, therefore it is necessary to have python installed in the system.
Documentation requirements
These are the required documents:
Requirements: Statements that identify attributes, capabilities, characteristics, or qualities of a product.
Architecture/Design: Overview of software, includes relations to an environment and construction principles to be used in design of software components.
Technical: Documentation of installation, code, algorithms, interfaces, and APIs.
Training: Documentation of training guide includes all steps of learning to implementation of the program.
