ITECH1103 Big Data And Analytics

0 Download 9 Pages / 2,077 Words

23-11-2020

Course Code: ITECH1103
University: Federation University

MyAssignmentHelp.com is not sponsored or endorsed by this college or university
Country: Australia

Question:

Data set background:

The Solar cities project was a project led by the University of Ballarat, (former name of Federation University), which involved the recruitment of households and businesses across the Loddon Mallee and Grampians regions to monitor changes in energy consumption. The project looked at a number of factors that could influence energy consumption. These factors were broken up into sets of features, and measurements were taken for each respective feature. For example, a factor could be related to a dwelling's construction materials. In which case a feature could be “dwelling construction type” and a measurement would be taken to determine the construction type for each dwelling and stored in the data set. For example “dwelling construction type” could contain the values brick, brick veneer etc... Many of these features are included within the provided Solar Cities data set.

The following are sets of features included in the provided data set:

Adoption of solar energy technologies

Geographic characteristics

Physical characteristics of the dwellings, including such things as the dwellings age, size, number of stories, number of lights, insulation etc.

The main aim of this project is to understand the drivers of power consumption, and as a large percentage of electrical energy is created by coal fired plants, then conversely the drivers of CO₂

Therefore two main questions that need to be asked as a researcher of the Solar Cities project are:

Which combination of features highlight where efficiencies could be made in the reduction in energy consumption?
What would you include in a predictive model that would explain the demand on future energy use and CO₂emissions?

i.e. which features and their measurements contribute to either higher or lower energy consumption? (For a predictive model, Watson Analytics does have a rule based engine...)

You have been commissioned as a research assistant with the main aim to find the drivers of power usage, understand the factors that contribute to this and provide any useful and interesting insights, trends and patterns that could be presented to the stakeholders of this project. Your intended audience are the project manager and associated stakeholders.

Your findings will be presented to the project manager and the other stakeholders, where you will outline the factors that contribute to power usage and what your recommendations will be to reduce power consumption, given the current dwellings that exist in the given geographical locations in country Victoria (i.e found in the data set).

You are expected to present the data findings in visual forms (i.e., charts and graphs).

Some Starting Questions

What is the contribution of power usage over a year by roof colour?

What is the contribution of power usage over a year by PV_Capacity?

What is the contribution of power usage over a year by PV_Capacity and Insulation?

What is the power usage by estimated age?

Over which months is the most power used?

Over which months is the least power used?

What are the top drivers of power usage?

Which suburbs have the most houses with pv_capacity

Which age houses are more likely to have pv_capacity?

Are houses that are owned more likely to use less power than the ones that are rented?

Which suburb dwellings use the most power?

Do houses with larger square meterage use more power than smaller houses, also does double story make a difference?

Which light types in dwellings use more power?

Does having more lights of any type mean the house will use more power?

What age houses have what type of wall construction?

What age houses and from which areas and with how many bedrooms use the most power?

Does Roof colour and roof material make difference to power consumption?

Do dwellings that have double glazed windows and with window coverings use less power?

Task 1- Background information

Write a description of the selected dataset and its importance given the Solar Cities project, i.e. the Solar Cities project is attempting to understand power usage in dwellings, in order to then recommend how to improve the overall efficiency of power consumption in dwellings. Information must be appropriately referenced.

Task 2 – Reporting / Dashboards

For your project, perform the relevant data analysis tasks by

answering the above questions

finding new relevant questions

answering the two main questions

You should also identify the visualization and dashboards you need to develop to best communicate your findings to the stakeholders.

Task 3 – Research

Justify why these BI reporting solution/dashboards are chosen in Task 2 (Reporting / Dashboards) and why those data set features are present and laid out in the fashion you proposed (feel free to include all other relevant justifications).

Note: To ensure that you discuss this task properly, you must include visual samples of the reports you produce (i.e. the screenshots of the BI report/dashboard must be presented and explained in the written report; use ‘Snipping tool’), and also include any assumptions that you may have made about the analysis in your Task2.

Task4 – Recommendations for the project manager and project stake holders

Based on your BI analysis and the insights gained from the “Data Set”, given your analysis performed in previous tasks, make some logical recommendations to the stake holders. Justify the answers to the two main questions, and show where power consumption efficiencies could be made, and what future consumption will look like. Give 2 possible scenarios. Also, if you have a predictive model show which features and measurements lead to power reductions/increases, to add to your logical recommendations. Do this with the help of appropriate references from peer-reviewed sources.

Task 5 - The Reflection: Each Team member is expected to write a brief reflection about this project in terms of challenges, learning and contribution.

Answer:

Background:

The data analysis of this study finds the “Skills and Credibility” of the students about the fact how they are capable of using the online visualisation tools- “IBM Weston Analytics” (Zhu et al., 2014). Also, their ability to execute the decision analysis and predictive analysis are verified in this research report. The research report is structured on the basis of “Solar city project data” analytically. The data set used in the report is gathered by a reputed concerning University. This analytical study focuses on the explanatory variables of the data set that may control the usage of electric power in various suburbs of “Victoria”. The analyst concluded as per the preferences of “Dwellings” and “Businesses” over “London Mallee” and “Grampians”.

The research questions are represented by “Dashboard” presentation. The three types of features considered here, are – 1) “Geographic Characteristics”, 2) “Adoption of Solar energy Technologies”, and 3) “Physical Characteristics of different types of dwellings”. The explanatory factors associated to the dwellings detain the significant drivers and predictors of the energy usage. A significant proportion of “Electric energy” generated from “Non-renewable” resources of energy especially “Coal” or “Petroleum” generate extensive amount of threatening gases like CO₂. The reporting of “Business analysis” with the support of “Dashboards” would support the decision-makers and policy-makers to commence the interpretations with respect to “Business outputs” (Jiang, 2017). The analyst would present the outcomes of the analysis to the project manager and stakeholders. They would be capable of recommending the possible steps and measures to make suggestive decisions regarding geographical locations of “Victoria” in the matter of excess energy usage and excess emission of carbon di-Oxide (Kang & Han, 2008).

Discussion:

“What is the contribution of power usage over a year by roof colour?”

The power was mostly used in 2014 and least in 2012. The power usage is highest in “Dark” coloured roof followed by “Intermediate” coloured roof.

“What is the contribution of power usage over a year by PV_Capacity?”

The power usage is highest when Photo-Voltaic capacity is least. The power is mostly used in the dwellings in 2014 for the Photo-Voltaic capacity 0 to 960. The overall power usage is least in Photo-Voltaic capacity 3841-4800.

“What is the contribution of power usage over a year by PV_Capacity and Insulation?”

The power usage is highest for the Photo-Voltaic capacity 0 to 960 with insulation 1 in 2014 followed by Photo-Voltaic capacity 0 to 960 with insulation 1 in 2013 and 2015. The power usage is comparatively lesser in the dwellings for higher Photo-Voltaic capacity and number of insulation does not put a significant effect in this case. The

“What is the power usage by estimated age?”

The usage of power is highest for the dwellings where estimated age of the dwellings is Sixty years and over. The usage of power is least for the dwellings whose estimated age is 0 to 4 years.

“Over which months is the most power used?”

The total power usage in all the four years is highest in “July” with the usage of more than 155 K.

“Over which months is the least power used?”

The total power usage in all the four years is lowest in “November” with the usage of less than 80 K.

“What are the top drivers of power usage?”

The significant “Top driver” of the power usage are –

“Suburb” singularly explain 21% consumption of the Power. On the other hand, “Wall construction” and “Suburb” both jointly explain 28% consumption of Power.

“Which suburbs have the most houses with pv_capacity?”

The “Portland” suburb have most number of houses with Photo-Voltaic capacity. The “Heywood” suburb have second highest number of houses with Photo-Voltaic capacity.

“Which age houses are more likely to have pv_capacity?”

The “Age group” of “Thirty to Thirty-Nine” are more likely to have a Photo-Voltaic capacity. The Power Voltic Capacity in this “Age group” varies from 0 and 3840.

“Are houses that are owned more likely to use less power than the ones that are rented?”

No, the “Owned” houses are more likely to use power than the “Rented” houses.

“Which suburb dwellings use the most power?”

The usage of power in all the dwellings is maximum for “Portland” suburb. The usage of power is second highest in “Heywood” suburb.

“Do houses with larger square meterage use more power than smaller houses, also does double story make a difference?”

The power usage is higher for the houses that has larger area (199 sq. metre) and power usage is lesser for the rooms whose area is small. The usage of power usage is higher for single storied houses than double storied houses.

“Which light types in dwellings use more power?”

From the five graphs of average power usage and light counts of LED, CFL, Halogen, Incandescent and Fluorescent lights, it is observed that “Halogen” light mostly causes power consumption.

“Does having more lights of any type mean the house will use more power?”

The power consumption is higher for the houses with any means of lights (“CFL”, “Halogen”, “LED”, “and Fluorescent” and “Incandescent”) when total number of lights is 17. The power consumption descends for the power usage 13, 21, 23 and 26. No prominent pattern is found for the power usage and number of lights. Hence, more number of lights do not mean the more amount of power usage.

“What age houses have what type of wall construction?”

The power usage is maximum for the houses made of “Weatherboard” having age Sixty years and over. The houses made of “Bricks” have second highest power usage for the age Twenty to Twenty-nine. Overall, the power is mostly used in the houses made of “Bricks” followed by “Weatherboard”. The power usage is least for the roof construction of “Concrete Block”, “Double Brick” and “Unknown” types of building materials.

“What age houses and from which areas and with how many bedrooms use the most power?”

The power usage is maximum in the houses of age “Forty to Sixty Nine” years with number of bedrooms 1 to 20 in “Portland” suburb.

“Does Roof colour and roof material make difference to power consumption?”

Yes, roof colour and roof material make difference to power consumption. The power is mostly consumed in “Dark” coloured rooms made of “Bricks”, followed by “Intermediate” coloured rooms made of “Bricks”. The power consumption is comparatively much lesser for the dwellings of “Concrete block” and “Double brick”.

“Do dwellings that have double glazed windows and with window coverings use less power?”

Yes, the dwellings having “Double glazed” windows and window coverings (Blinds or Curtains) utilise lesser power than the dwellings of “Single glaze” windows with and without any kinds of “Window coverings”.

Dashboards:

In case of “Business intelligence” and “Business information”, the reporting by dashboard is a vital aspect that assistances to make conversant conclusions and decisions especially for the “IP Professionals” (Eckerson, 2010). Decision-makers and policy makers uses “Dashboards” to capture the “Long-term Objectives” (Ferrucci, 2012). “Creating”, “Managing” and “Sharing” of the business reports could be an activity represented by “Dashboards” for its versatility (Palpanas et al., 2007). This analysis is visualized with the help of “Dashboards” that easily reveals “Inherent trend”, “Pattern” and “Association” among the factors of the data set.

Research Question 1.

“Which combination of features highlight where efficiencies could be made in the reduction in energy consumption?”

Recommendations to reduce power usage:

The combinations that would be beneficial to reduce the power consumption are-

Double Glazed window without window coverings.
The newly constructed dwellings with estimated age less than 5 years.
Rooms with “Light” colour with wall-construction type “Timber” and “Fibro” built.
The double storied dwellings with small size rooms.

If following aspects are considered at the time of constructing the dwellings, then lesser power would be consumed (Hoyt et al., 2016).

Research Question 2.

“What would you include in a predictive model that would explain the demand on future energy use and CO₂ emissions?”

Recommendations to predict power usage:

The dashboard presents the predictive model of use of energy consumption by different factors that would determine the emission of poisonous gas like Carbon-di-Oxide (CO₂). These factors are –

The power usage is gradually growing year after year from 2012 to 2014; however, the consumption dropped in last year (2015).
The total power usage is high in “Owned” house rather than “Rented” house.
Lesser number of bedrooms consume lesser amount of power in terms of electricity.
The key of power usage is “Wall-Construction” and “Suburb” simultaneously with explanatory power 28%.
The predictive strength to estimate “Power usage” by “Suburb” and 14 other explanatory factors is 52%.
The “dark” colour rooms causes more power usage than other coloured rooms.
The most significant predictors are “Suburb”.

Research Reporting:

The two constructed dashboards indicate that dark coloured rooms, with high estimated age, brick type wall constructing material, single storied dwellings and with window curtains consumes high amount of energy. The lower PV_capacity and heavy usage of lights especially “Halogen” lights consumes high amount of electric energy. As the year is proceeding, the power consumption in the different types of suburbs especially in “Heywood” and “Portland” suburb is significantly enhancing. Size of the rooms and different types of month of the year also bring variability of power usage.

Reflections:

The research study reveals the causes and their significance for the consumption of conventional electric-energy in various suburbs of “Victoria” (Frolick & Ariyachandra, 2006).
The analysis helped the analyst to learn the procedure of handling big data (Lim, Chen & Chen, 2013).
However, the presence of “Unknown” values in several factors of the data set is quite confusing and disturbing.
The online analytic tool “IBM Watson Analytics” is successful to handle data sets and data visualization with proper graphs and tables.
The dashboards helped to find the probable recommendations from the end of a “Project manager” and “Stakeholders”. It would make the research study fruitful (Chen, Chiang & Storey, 2012).

References:

Chen, H., Chiang, R. H., & Storey, V. C. (2012). Business intelligence and analytics: from big data to big impact. MIS quarterly, 1165-1188.

Eckerson, W. W. (2010). Performance dashboards: measuring, monitoring, and managing your business. John Wiley & Sons.

Ferrucci, D. A. (2012). Introduction to “this is watson”. IBM Journal of Research and Development, 56(3.4), 1-1.

Frolick, M. N., & Ariyachandra, T. R. (2006). Business performance management: One truth. IS Management, 23(1), 41-48.

Hoyt, R. E., Snider, D., Thompson, C., & Mantravadi, S. (2016). IBM Watson analytics: automating visualization, descriptive, and predictive statistics. JMIR public health and surveillance, 2(2).

Jiang, F. (2017). Data Analytics Helps Business Decision Making.

Kang, J. G., & Han, K. H. (2008, November). A business activity monitoring system supporting real-time business performance management. In Convergence and Hybrid Information Technology, 2008. ICCIT'08. Third International Conference on (Vol. 1, pp. 473-478). IEEE.

Lim, E. P., Chen, H., & Chen, G. (2013). Business intelligence and analytics: Research directions. ACM Transactions on Management Information Systems (TMIS), 3(4), 17.

Palpanas, T., Chowdhary, P., Mihaila, G., & Pinel, F. (2007). Integrated model-driven dashboard development. Information Systems Frontiers, 9(2-3), 195-208.

Zhu, W. D. J., Foyle, B., Gagné, D., Gupta, V., Magdalen, J., Mundi, A. S., ... & Triska, M. (2014). IBM Watson Content Analytics: Discovering Actionable Insight from Your Content. IBM Redbooks.