Term Deposit is considered as a cash investment

Answer to question [b]

Refer to the exploratory analysis of both the data set as mentioned in table 1 and table 2, it can be said that three missing variables are there in case of bank_train.csv data set [contact, month and poutcome], whereas the data set bank_test.csv also contained three missing variables [contact, month and poutcome]. This indicates that there are no unusual patterns evidenced in both the data set.   However, it is noted from the below mentioned histograms that there is a potential relationship between predictor variables and target variable outcome. Mainly, age, balance, duration and pdays have the major influence on target outcome variable.

Answer to question [c]

Decision tree analysis

Answer to question [d]

Logistic regression

In order to conduct the logistic regression model “y” is considered as the predictor variable and age, marital, education, default, balance, housing, loan, duration, campaign and pdays are considered as the target variable. With the use of rapid miner tool, at first the below mentioned model has been established. Then the logistic regression analysis has been performed.

 

Exploratory analysis of bank_train.csv data set
Role	Table Index	Name	Type	Statistics	Range	Missing
Regular	0	Att1	Integer	Avg = 22602.848 +/- 13082.182	[1.000 ; 45209.000]	0.0
Regular	1	Age	Integer	Avg = 40.931 +/- 10.608	[18.000 ; 95.000]	0.0
Regular	2	Job	Polynominal	Mode = blue-collar (7367), least = unknown (200)	Management (7092), technician (5693), entrepreneur (1116), blue-collar (7367), unknown (200), retired (1712), admin. (3837), services (3083), self-employed (1189), unemployed (971), student (715), housemaid (934)	0.0
Regular	3	Marital	Polynomial	Mode = married (20352), least = divorced (3892)	Married (20352), single (9665), divorced (3892)	0.0
Regular	4	Education	Polynomial	Mode = secondary (17387), least = unknown (1360)	Tertiary (9989), secondary (17387), unknown (1360), primary (5173)	0.0
Regular	5	Default	Binominal	Mode = no (33288), least = yes (621)	No (33288), yes (621)	0.0
Regular	6	Balance	Integer	Avg = 1362.096 +/- 3090.431	[-8019.000 ; 102127.000]	0.0
Regular	7	Housing	Binominal	Mode = yes (18865), least = no (15044)	Yes (18865), no (15044)	0.0
Regular	8	Loan	Binominal	Mode = no (28483), least = yes (5426)	No (28483), yes (5426)	0.0
Regular	9	Contact	Binominal	Mode = cellular (21935), least = unknown (9802)	Unknown (9802), cellular (21935)	2172.0
Regular	10	Day	Integer	Avg = 15.785 +/- 8.332	[1.000 ; 31.000]	0.0
Regular	11	Month	Binominal	Mode = may (10386), least = jun (4023)	May (10386), jun (4023)	19500.0
Regular	12	Duration	Integer	Avg = 257.853 +/- 256.930	[0.000 ; 3881.000]	0.0
Regular	13	Campaign	Integer	Avg = 2.757 +/- 3.073	[1.000 ; 63.000]	0.0
Regular	14	Pdays	Integer	Avg = 40.072 +/- 100.121	[-1.000 ; 854.000]	0.0
Regular	15	Previous	Integer	Avg = 0.577 +/- 2.417	[0.000 ; 275.000]	0.0
Regular	16	Poutcome	Binominal	Mode = unknown (27752), least = failure (3648)	Unknown (27752), failure (3648)	2509.0
Regular	17	Y	Binominal	Mode = no (29942), least = yes (3967)	No (29942), yes (3967)	0.0

Table 1

 

Exploratory analysis of bank_test.csv data set
Role	Table Index	Name	Type	Statistics	Range	Missing
regular	0	att1	integer	avg = 22615.456 +/- 12959.324	[12.000 ; 45211.000]	0.0
regular	1	age	integer	avg = 40.953 +/- 10.650	[18.000 ; 94.000]	0.0
regular	2	job	polynominal	mode = management (2366), least = unknown (88)	admin. (1334), entrepreneur (371), technician (1904), services (1071), blue-collar (2365), management (2366), retired (552), housemaid (306), unemployed (332), self-employed (390), student (223), unknown (88)	0.0
regular	3	marital	polynominal	mode = married (6862), least = divorced (1315)	single (3125), married (6862), divorced (1315)	0.0
regular	4	education	polynominal	mode = secondary (5815), least = unknown (497)	secondary (5815), unknown (497), primary (1678), tertiary (3312)	0.0
regular	5	default	binominal	mode = no (11108), least = yes (194)	no (11108), yes (194)	0.0
regular	6	balance	integer	avg = 1362.800 +/- 2903.585	[-6847.000 ; 71188.000]	0.0
regular	7	housing	binominal	mode = yes (6265), least = no (5037)	yes (6265), no (5037)	0.0
regular	8	loan	binominal	mode = no (9484), least = yes (1818)	no (9484), yes (1818)	0.0
regular	9	contact	binominal	mode = unknown (3218), least = telephone (734)	unknown (3218), telephone (734)	7350

 

regular	10	day	integer	avg = 15.872 +/- 8.295	[1.000 ; 31.000]	0.0
regular	11	month	binominal	mode = may (3380), least = jun (1318)	may (3380), jun (1318)	6604.0
regular	12	duration	integer	avg = 259.095 +/- 259.323	[3.000 ; 4918.000]	0.0
regular	13	campaign	integer	avg = 2.783 +/- 3.171	[1.000 ; 51.000]	0.0
regular	14	pdays	integer	avg = 40.576 +/- 100.154	[-1.000 ; 871.000]	0.0
regular	15	previous	integer	avg = 0.589 +/- 1.925	[0.000 ; 51.000]	0.0
regular	16	poutcome	binominal	mode = unknown (9207), least = failure (1253)	unknown (9207), failure (1253)	842.0

Task 2:

 

The process of datafication is rapidly increased especially in the environment of corporate sector. Big data and associated phenomena is the current trend for keeping the information a server and execute it in terms of making decision for further success (Zwitter, 2014). One of the most effective phenomena of the datafication for business or individual is social media that suppressed the capacity for changing average consumer in terms of understanding benefits of the company (Krasnow Waterman and Bruening, 2014).

This particular study provides the information that relates with the business ethics that has to be perceived. However, in this particular topic analyst explain how the big data analytic tools has effective impact on the decision making process for the company and its impact on the ethical conceptions. 

Traditional Ethics

According to Hendler (2014), from the point of view of moral responsibility of the business as well as individual, utilitarian and deontological tradition has places their strong emphasize. Therefore, for the individual, these moral agency has the much stem that regularly followed the assumptions for the busies that helps in making decision (Krasnow Waterman and Bruening, 2014). However, Hill (2014) argued that as the world goes changes and modern technologies implement in the business process, all the assumption changes and faced several challenges especially the big data. The moral agency determinate that the degree of the entity process is the key responsible objects of that entity (Thimmarayappa and Voges 2014). In combination with the intrinsic factors and extraneous, moral responsibility escape the entity as well as it defines the culpability. Following are the several entity that innate the conditions such as –

Knowledge: The agent of the big data analytic tools is able blame for the result of the actions that relates with the business process as well as analytics (Hoskins, 2014). Apart from that, agent can be blamed for the knowledge of the consequence regarding the actions that already taken for big data analyses in terms of making decision.

Causality: Bholat (2015) argued that agent is able to hold the responsibilities for the ethical events that relevant to the business results especially the outcomes of the actions that taken for decision making and future improvement in big data analytics environment.

Choice: If the business has liberty for choosing the alternative without any harm in the business process, agent is able to blame for the result (Schneider, 2011).

Apart from that, the agents are also able to observe the tend that exculpate the process full moral agency. This process is executed if the above mentioned three different options does not execute (Krasnow Waterman and Bruening, 2014). There are several considerations those morally relevant outcomes of the business organization from the point of view of big data analytics (Kirkpatrick, 2013). The sense of negative consequence is also helps in establishing the moral obligations.

Ethical big data challenges

Big data has several qualities such as it helps in collecting several options that determinate the time taking for decision making, impose the new rules and regulations and create the innovation as well as knowledge through developing the bridge with the new datasheets and old datasheets (Panchadsaram, 2014). Apart from that, big data allows the business organization as well as individual in making successful decision through analyzing the entire activities (Krasnow Waterman and Bruening, 2014). All the business organizations have to implement the rules and regulations of the data analytics to carry forward their business in an ethical manner.  All the given rules and regulations assists the organization to implement the tool of business analytics in an effective manner.

However, Big data could be big (2011) argued that ethical challenges regarding big data analytic tools increased day to day such as global warming effect on the emissions of the many business organizations and individuals. Citraro (2013) opined that big data analytics is the possible effect of the individuals as well as sensory data for business. Moreover, Correction (2013) acknowledged that in terms of measuring the real data throughout the world big data creates lots of complexity and digitalized image for the business. Measurement of the digital image creation is specially calls the datafication (Gaff, Sussman and Geetter, 2014). It has been simply identified that absence of the knowledge affect on the data regarding data generators such as cell phone owning people, online consumers (Leonard, 2013). The contribution of internet of things contribute lots of methods in terms of maintaining appropriate business environment between the actor of knowledge and actor source of power and information.

Couldry and Powell (2014) examined that leads of global data imbalance between the different stakeholders that mostly benefitted the corporate agencies in terms of generating knowledge of information and intelligence. Moreover, Dhar (2014) presumed that big data correlations suggested several causations that might be effect the ethical behavior especially when the organization implement the true Delphian oracle within their business process for decision making and keeping the business information in central server. It is underlying the vulnerability that having believes.

Privacy

The life in these days becomes mirror and majority of people become the aspects of cyber reality and recorded. Majority of the organization becomes completely transparent for the actors for their right access and skills (Lesk, 2013). In order to control the privacy in datafication, the Guardian developed the RIOT (Rapid Information Overlay Technology) that allows to track or access the information from the social network site including IP address (Gaff, Sussman and Geetter, 2014). It can allow to access information of one person to complete their everyday duty.

Group Privacy

According to Dumbill (2013), majority of data analysts use big data in terms of accessing information in terms of finding out the shopping preferences, sleep cycles, online consumptions, health status, etc. Moreover, there are some data analysts who used big data in terms of accessing information that are most intelligence from the point of view of individualization and de-individualization. This is the form of one aspect of anonymization (Lyon, 2014). However, regarding the group privacy, issues are already identified. The sense of de-individualization is always become the more transparent from the point of view of statistical database. However, as the statistical data are increasing rapidly, big data is used to analyze in a shortest and easiest way. Hand (2011) cited that big data is used in a certain way that consume the information about the behavior of the people. Therefore, in the different way, it can be employed for encouraging as well as discouraging the certain behavior of the people. Moreover, through using big data, data analyst is able to discover the hidden correlations and increase the effectiveness of data (Neff, 2013). Big data has the ability for creating incentives in terms of minimizing the less transparent information within the database. Hayes (2011) utilized that hyper connectivity of the big data is also allowed the data analysts in implementing other strategies such as bots in the infiltrate twitter.

Ethics

In terms of applying the big data analytics tools, organisation has to maintain several information including standards and codes that lay behind the development of data analysis method with proper analysis. Majorly it has been identified that ethics of research concern the privacy questions regarding use of social media such as Facebook, Google Plus, Twitter, etc that are the usual suspect from the point of view of privacy. As argued by Dumbill (2013), Facebook is the key usual suspect that comes first regarding the question of privacy in datafication. Therefore, data analytics tools especially the big data analytics tool has lot of reveal that much about the very specific group in the geographical relations. On the other hand, Eckhoff and Sommer (2014) opined that individual information is not working properly in terms of investigating the purpose of intelligence. For the individual tag, valuable information for the companies does not required. The main problem that raised regarding this information is group privacy in big data analytics. Majority of organization does not maintain or follow the ethical rules as well as codes in terms of mitigating the non-privacy related ethical effect such as preserving of integrity, BD&S own statement, privacy of subject participation, etc. Inform consent is the best ethics that despite the data from becoming published. Gaff, Sussman and Geetter (2014) argued that in terms of analyzing and representing the pertinent social phenomena, data analysts collected information from the social media such as Twitter, LinkedIn, Facebook, etc without considering the lack of informed consent. According to the constitution, this the major breach for research ethics.

Legislation

Various legislations are required for Big Data analytics in case of every business organizations. All the business organizations are required to comply with the rules and regulations in terms of big data analytics (Gosain and Chugh, 2014).  The arts and humanities research council has laid down several rules and regulations in terms of legislation requirements and amendment structure of the organization.  The various other legislations include Privacy Act, 1988, Crimes Act, 1914, Archives Act 1983, Evidence Act 1995 and Public Governance Act and Accountable act, 2012.

Task 3

1. Best Performing Countries

It can be seen that Nigeria is the top performing nation among the top 10 list with 127246092 mobile phone adoption in 2013. Nigeria is followed by Ghana, Mali, Cote d’Lvoire, Senegal, Niger, Mauritania, Liberia, Gambia and Guinea-Bissau.

Nations (top 10)	Adoption of Mobile Phones (2013)
Nigeria	127246092
Ghana	28026482
Mali	19749371
Cote d’Lvoire	19390902
Senegal	13133772
Niger	7006300
Mauritania	4025478
Liberia	2550775
Gambia	1848854
Guinea-Bissau	1262700

 

2. Worst Performing Countries

Ghana is the worst performing nation in adopting internet technology. Around 3502202 people does not use internet. Ghana is followed by Cote d’Lvoire, Mali, Congo Repulic, Gambia, Liberia, Guinea, Sierra Leone, Gunia-Bissau and South Sudan.

Nations (worst 10)	Adoption of Internet (2013)
Ghana	3502202
Cote d’Lvoire	528218
Mali	351938
Congo Repulic	293544
Gambia	258900
Liberia	197528
Guinea	187923
Sierra Leone	103565
Gunia-Bissau	52832
South Sudan	0

 

3. Best Performing countries in terms of adoption of landlines technology

Cabo Verde is the best performing nation in comparison to other nine nations. Since 2000, Cabo Verde is topping the list in Africa. 0.1326 per head population is adopting the landlines technology. Cabo Verde is followed by Gambia, Senegal, Mauritania, Cote d’Lvoire,  Ghana, Mali, Niger, Guinea-Bissau and Nigeria.

 

Nations (best 10)	Adoption of Landline (2013)
Cabo Verde	0.1326
Gambia	0.0347
Senegal	0.0243
Mauritania	0.0139
Cote d’Lvoire	0.0134
Ghana	0.0104
Mali	0.0083
Niger	0.0056
Guinea-Bissau	0.0029
Nigeria	0.0021

 

4. Summary of Key Technology Adoption

The adopters of technology have significantly increased from 2000 to 2013. The adopters of mobiles were 42.97, internet was 7.878 and landlines were 1.859 in 2013.

Year	Mobile/Population	Internet/Population	Landlines/Population
2000	1.36	0.449	1.657
2001	2.16	0.619	1.762
2002	3.01	0.870	1.831
2003	4.03	1.112	1.906
2004	5.63	1.626	1.945
2005	8.29	1.873	1.968
2006	11.58	2.292	1.966
2007	15.88	2.803	2.065
2008	21.01	3.478	2.081
2009	25.52	3.734	2.169
2010	30.82	5.366	2.126
2011	35.24	6.315	2.135
2012	39.02	6.943	1.949
2013	42.97	7.878	1.859

 

3. Rationale for Graphic Design and Functionality

1. Best Performing Countries

For this section, the Y-line has been set as mobile and X-line has been set has year. For each nation, the data regarding adoption of mobiles from year 2000 to 2013 has been mentioned. Therefore, it helps in knowing in which year the population of particular nation has adopted mobile technology which is pretty clear that Nigeria has topped the list by huge margin. In the graph, top section represents 10 countries of Africa for assessing the nations with most adoption of mobile technology.

2. Worst Performing Countries

In this part, y-line represents adoption of internet by population and top side represents the top ten worst performing nations in adopting the technology. On the other hand, x-axis has been taken as year representation. The nations have been segmented on the basis of year from 2000-2013 for analyzing the adoption of internet technology. Thus, it can be acknowledged from the graph that Ghana is in top whereas South Sudan is at below. Apart from that, value of adoption of technology is easily known for each year so that better utilization can be made such as in decision making or making strategy to increase the adoption of technology among the population.

3. Best Performing countries in terms of adoption of landlines technology

In this section, Y-axis represents adoption of landline by the population whereas X-axis shows years from 2000-2013. Again, top axis represents nations of Africa which are ten best performing nation in using the landline technology by per head of population. Thus, the design helps in showing the data of each nation from 2000 to 2013 in adopting technology. For every nation of Africa, data of region, country name and year I represented which would provide great assistance to the users of data and using it for making decision and making further analysis. It can be make out from the graphic design of this section is that Cabo Verde is topping the list in all consecutive years from 2000 to 2013. Apart from that, use of technology in Niger has significantly gone down. Thus, it can be mentioned that graphic design clearly depicts that increase and decline in the adoption of technology can be effectively known and useful analysis can be drawn for gaining better results.

4. Summary of Key Technology Adoption

The graphic design of this section is divided into three technologies such as mobile, internet and landlines. These technologies have been drawn in context to population of regions/nations. The details of all technologies are easily represented for each year that is from 2000 to 2013. Thus, it helped in knowing how much the value have increased or decreased from year to year. It is understood that adoption of mobile technology is higher in comparison to other technology that is internet and landlines by the population of African’s nation. Therefore, it can be drawn that designing an effective graphic design always showcase a better platform for making effective analysis.

Reference List

 

Bholat, D. (2015). Big Data and central banks. Big Data & Society, 2(1).

Big data could be big. (2011). Int J Productivity & Perf Mgmt, 60(8).

Citraro, D. (2013). On Visualization. Big Data, 1(1), pp.14-17.

Correction. (2013). Big Data, 1(2), pp.113-113.

Couldry, N. and Powell, A. (2014). Big Data from the bottom up. Big Data & Society, 1(2).

Dhar, V. (2014). Why Big Data = Big Deal. Big Data, 2(2), pp.55-56.

Dumbill, E. (2013). Big Data is Rocket Fuel. Big Data, 1(2), pp.71-72.

Dumbill, E. (2013). Making Sense of Big Data. Big Data, 1(1), pp.1-2.

Eckhoff, D. and Sommer, C. (2014). Driving for Big Data? Privacy Concerns in Vehicular Networking.IEEE Secur. Privacy, 12(1), pp.77-79.

Gaff, B., Sussman, H. and Geetter, J. (2014). Privacy and Big Data. Computer, 47(6), pp.7-9.

Gosain, A. and Chugh, N. (2014). Privacy Preservation in Big Data. International Journal of Computer Applications, 100(17), pp.44-47.

Hand, E. (2011). Google research chief pushes 'big data'. Nature.

Hayes, J. (2011). Data centres: the next big challenges. Engineering & Technology, 6(9), pp.50-53.

Hendler, J. (2014). Data Integration for Heterogenous Datasets. Big Data, 2(4), pp.205-215.

Hill, S. (2014). TV Audience Measurement with Big Data. Big Data, 2(2), pp.76-86.

Hoskins, M. (2014). Big Data 2.0: Cataclysm or Catalyst?. Big Data, 2(1), pp.5-6.

Kirkpatrick, R. (2013). Big Data for Development. Big Data, 1(1), pp.3-4.

Krasnow Waterman, K. and Bruening, P. (2014). Big Data analytics: risks and responsibilities.International Data Privacy Law, 4(2), pp.89-95.

Leonard, P. (2013). Customer data analytics: privacy settings for 'Big Data' business. International Data Privacy Law, 4(1), pp.53-68.

Lesk, M. (2013). Big Data, Big Brother, Big Money. IEEE Secur. Privacy, 11(4), pp.85-89.

Lyon, D. (2014). Surveillance, Snowden, and Big Data: Capacities, consequences, critique. Big Data & Society, 1(2).

Neff, G. (2013). Why Big Data Won't Cure Us. Big Data, 1(3), pp.117-123.

Panchadsaram, R. (2014). Untapped Potential. Big Data, 2(2), pp.63-64.

SCHNEIDER, M. (2011). Data Show Big Drop in Central Line Infections. Hospitalist News, 4(3), p.16.

Thimmarayappa, S. and V, M. (2014). Big Data Privacy and Management. International Journal of Computer Applications, 107(6), pp.13-16.

Zwitter, A. (2014). Big Data ethics. Big Data & Society, 1(2).

Rotberg, R. I., and Aker, J. C. (2013). Mobile phones: uplifting weak and failed states. The Washington Quarterly, 36(1), 111-125.

Carmody, P. (2013). A knowledge economy or an information society in Africa? Thintegration and the mobile phone revolution. Information Technology for Development, 19(1), 24-39.

Braun, R., Catalani, C., Wimbush, J., and Israelski, D. (2013). Community health workers and mobile technology: a systematic review of the literature. PloS one,8(6), e65772.

Ngimwa, P., and Wilson, T. (2012). An empirical investigation of the emergent issues around OER adoption in Sub-Saharan Africa. Learning, Media and Technology, 37(4), 398-413.

Pamuk, H., Bulte, E., and Adekunle, A. A. (2014). Do decentralized innovation systems promote agricultural technology adoption? Experimental evidence from Africa. Food Policy, 44, 227-236.

Tang, C. Y., Lai, C. C., Law, C. W., Liew, M. C., and Phua, V. V. (2014). Examining key determinants of mobile wallet adoption intention in Malaysia: an empirical study using the unified theory of acceptance and use of technology 2 model. International Journal of Modelling in Operations Management, 4(3-4), 248-265.

Hanafizadeh, P., Keating, B. W., and Khedmatgozar, H. R. (2014). A systematic review of Internet banking adoption. Telematics and informatics, 31(3), 492-510.

Sahd, C., and Thinyane, H. (2012). Community Telephone Networks in Africa-Bridging the Gap Between Poverty and Technology. International Journal on Advances in Networks and Services, 5(1 and 2), 129-138.

Shambare, R. (2013). Factors influencing the adoption of cell phone banking by South African students. African Journal of Business Management, 7(1), 30.

Brown, M., and Stix, H. (2014). The euroization of bank deposits in Eastern Europe. Economic Policy, eiu002.

Giannone, D., Lenza, M., and Reichlin, L. (2012). Money, credit, monetary policy and the business cycle in the euro area

Gertler, M., Kiyotaki, N., and Queralto, A. (2012). Financial crises, bank risk exposure and government financial policy. Journal of Monetary Economics, 59, S17-S34.