Note you must use RapidMiner Studio for Task 2 and you must use Tableau Desktop for Task 3 in this Assignment.

63967 - Solution
WRITTEN PRACTICAL REPORT
TABLE OF CONTENTS:
1. DATA ANALYTICS & DATA WAREHOUSING TECHNIQUES
1.1 PREDICTIVE ANALYSIS
1.2 DATA WAREHOUSING STRUCTURES
1.3 KEY FACTORS & BEST PRACTICES
2. EXPLORATORY DATA ANALYSIS & LINEAR REGRESSION ANALYSIS
2.1 (i) EXPLORATORY DATA ANALYSIS
2.1 (ii) RESULTS OF EXPLORATORY DATA ANALYSIS
2.1 (iii) DISCUSSION & CONCLUSION
2.2 (i) LINEAR REGRESSION MODEL PROCESS
2.2 (ii) RESULTS OF FINAL REGRESSION MODEL
2.2 (iii) DISCUSSION & CONCLUSION
3. TABLEAU DESKTOP VIEW OF HOUSING DATA
3.1 BAR GRAPH
3.2 SCATTER PLOT
4. LIST OF REFERENCES
5. APPENDICES
TASK 1 - DATA ANALYTICS & DATA WAREHOUSING TECHNIQUES
1.1 PREDICTIVE ANALYSIS -
It is a branch of analytics which use data mining, statistics, machine learning and other modelling techniques to analyse the data so as to predict the future events based on the pattern in the current data.
With the use of predictive analysis, a firm can take decision that are scientifically proven rather than based on blind assumptions. Now a days it is being applied in almost every field like - healthcare, finance, sports, retail, cybersecurity and etc.
Techniques like data mining, statistics, text analytics, time series modelling help the user to create a relevant model by uncovering any hidden patterns in the large dataset in order to gain meaningful insights out of it, so as to predict the future values. The data can be structured and unstructured. Structured data does not need any cleaning process while unstructured data involves cleansing and various other transforming processes before using it for any model building process. (Aldowah, 2019)
EXAMPLE -
Here, we shall be considering an example of a firm which wants to undertake a promotional campaign in order to increase its revenue. Now it has three possible channels for promoting its product. Newspaper, YouTube and Facebook. A firm needs to decide the best channel option out of three to promote the product in order to have maximum effect on the sales. In the above case we can build a linear regression to get the coefficients associated with each promotional channel. This would help us to know by how much the sales increases by spending one unit of $USD on each of the promotional channels. And the eventually we can decide which would be the most effective cannel to promote our product.
1.2 DATA WAREHOUSING STRUCTURES -
A data warehouse can be termed as a central location where the information is collected & stored from multiple sources, so as to use that information in a meaningful way. Data warehouse architecture is the structure how a particular data warehouse is designed and its relationship with various other elements. (Warehousing, n.d.) (Warehousing, n.d.)
Based on the specific business needs, the data warehouse can be organised or architected in many different ways. The most popular five types are -
1) Hub and Spokes Architecture -
It is one of the most widely used architecture. In this architecture the data in operational system is moved from database into data warehouse staging area and then into the data warehouse and finally into the data Marts.
2) Stationary Data Warehouse -
In this architectural design the data is not changed from the source and the user has the direct access to the data.
3) Virtual Data Warehouse -
In this architecture the user has a direct access to the operational database by using any tool. This ensured flexibility aweel as minimises the redundancy of information.
4) Multi Stage Data Warehouse -
In this architectural design the Data Staging takes place multiple times before getting loaded into the data warehouse and finally into the respective Data Marts. It is suitable...