Predictive Analytics in Data Science: Applications and Challenges

Desklib provides solved assignments and past papers to help students succeed.

Predictive Analysis
Student name:
Student ID:
1

Table of Contents
Introduction......................................................................................................................................3
Body.................................................................................................................................................4
Conclusion.......................................................................................................................................9
Reference List................................................................................................................................10
2

Introduction
Predictive analytics can be defined as a set of diversified techniques of machine learning,
predictive modelling, and data mining. Predictive analytics aids to evaluate the historical as well
as current facts so that predictions can be made with respect to future or unknown events.
Predictive analytics is a branch of advanced analytics that is of great use for the organisations
such that it can achieve its objectives that have been set in varying conditions or situations.
Predictive analytics is of great importance for the organisations so that data predictions can be
made and predictive power can be assessed. Predictive analysis has six roles namely
measurement development, generation of new theories, assessing the forecasting of empirical
incident or phenomenon, evaluation of theories of competency, assessment of relevance, and
enhancement of the models that are existing.
In this report, the role played by predictive analytics have been described along with the
illustration of its application in present as well as future. The background of predictive analytics
has been described in the assignment that depicts the analysis of history of the role played by
predictive analysis. The role played by predictive analytics has been analysed in the report so
that its pros and cons can be identified.
3

Body
Background
The stream of information system has developed up to a huge extent in the last few years with
respect to the development of technique of statistical modelling such that empirical research can
be supported. In the present scenario, attention on formative constructs as well as selection bias
has increased as most of the IS (information system) researchers make use of modelling
corresponding to structural equation. However, it is required to make several improvements to
the same in order to enhance predictive analytics (Kessler, 2018). Predictive is of high
importance pertaining to the analysis of historical as well as current data so that the customers
can be comprehended and the patterns of purchases made by the customers can be
comprehended. It also aids to forecast the opportunities as well as dangers that are awaiting an
organisation.
The predictive analysis was commenced in 1940s, when the governments started to use early
computers. Analytics was first used by the organisations in the 19th century when Frederick
Winslow Taylor initiated exercises of time management. Another instance that can be related to
the use of analytics is by Henry Ford who calculated the pace of assembly lines. In the later part
of 1960s, analytics started receiving more attention as the computers were used as a support
system for making decisions (Shroff, 2017). The development of cloud, big data, and data
warehouses further stimulated the development of data analytics in order to make predictions
with respect to the operation of the organisations. Thus, predictive analytics evolved that is based
on data or statistics. It has been found that the construction of Pyramids in Egypt was based on
the use of statistics. Thus, in the present era, predictive analytics has gained huge importance for
organisations worldwide.
The functions of Predictive Analytics with respect to Scientific Research 498
Predictive analytics is very important with respect to the operation of the organisations. It has
been found that predictive analytics aids to develop as well as analyse theoretical models using a
variety of statistical models that are explanatory. Thus, predictive analytics plays a crucial role in
scientific research pertaining to models of explanatory statistics. The roles of predictive analysis
with respect to the role it plays in scientific research have been described below:
 Generation of a new theory:
4

The arguments that have been made by Strauss and Glaser with reference to grounded theory
states that in the case of theory building both qualitative as quantitative information can be used.
However, the authors have mainly emphasised on the use of quantitative statistics for the
creation of new models of prediction. Thus, it can be stated that predictive analytics plays a
crucial role in the generation of theories using statistics (Shmueli and Koppius, 2011).
 Development of measures:
Predictive analytics plays a crucial role in supporting the development of theory with respect to
constructing operationalisation. The development of new theories goes side by side of the
generation of new measures. Predictive analytics is several cases are used for the comparison of
various construct operationalisation taking into consideration the competence or capability of
users or different instruments of measurement.
 Assessing the forecasting of empirical incident or phenomenon:
Predictive analytics plays an important part in assessing the predictability of an event. It aids to
calculate the level or stage of predictability in terms of incidents that are measurable. The
knowledge pertaining to predictability is one of the core elements of scientific knowledge (Harris
et al., 2016). In case the level or stage of predictability is low then the creation of fresh measures
can be encouraged or spurred.
 Evaluation of theories of competency:
The models pertaining to explanatory statistics can be used as a medium of comparison or
evaluation. However, if the models are not formulated statistically then the comparison of the
theories cannot be compared. Predictive analytics whether predictive or explanatory, aids to
compare or evaluate models based on the examination of predictive accuracy such that the
competitiveness of the theories can be compared or evaluated (Parikh et al., 2016).
 Assessment of relevance:
Scientific growth or development requires rigorous investigation or research. Predictive analytics
can be used in order to assess the distance that exists between practice and theory. The power of
predictive analytics is enough to focus on the performance or presentation of experimental
model. Thus, predictive analytics can be used to analyse the significance of theories as well as
practices related to it.
5

 Enhancement of the models that exist:
Predictive analytics can aid to capture or obtain complicated relationships and patterns that are
underlying such that the statistical explanatory model that exists can be enhanced. Predictive
analytics can also aid to analyse the effect that investments made on IT have on the productivity
of organisations in hospitality industry (Shah et al., 2018). Predictive analytics has been used in
order to resolve the conclusions that are mixed with respect to the explanatory models that have
been explained earlier.
Empirical Models for Explanation
According to Collins (2018), the predictive analytics tend to help in devising the specific
strategies that would help in integrating the empirical methods providing the help and support in
carrying out the operations of the organisation by sticking to the policies and strategies. The term
explanatory statistical model is utilised to be used in a proper manner by sticking to the varied
range of policies and strategies of the Company. According to McCullagh (2019), explanatory
statistical model is utilised to explain the statistical model constructed to specify the reason of
causal hypotheses that provides description of the features of the statistical model utilised within
the analytics. Initiation is done from the level of theoretical models and the hypotheses are being
assessed by utilising the models and inference associated with the statistical methods and
strategies.
According to Alemi (2018), the explanatory statistical models tend to integrate the components
of explanatory statistical models that are associated with testing the causal hypotheses and
establishing the different activities and services of the Company in a successful manner by
sticking to the various policies. The processes involved in assessing the models of the
organisations tend to provide the basic steps and measures that help in establishing the
relationship.
Empirical Models for Prediction
According to Neuendorf (2016), the term and the concept of predictive analytics is being utilized
within the organizational sphere placed in the background that would help in establishing a
specific model to organize the varied range of activities and tasks related to the organizations.
Two components are involved within the organization that would help in assessing the predictive
6

models integrated with the empirical data and information. The empirical models for prediction
are specified within the organizational sphere to help in designing the newer or future
observations or features that are substantial.
The predictive power is related to the ability of the model to develop accurate and logical
interpretations of the operations and relevant services existing within the organizational sphere.
Empirical prediction is thought of as the empirical activity that involves the empirical goals to
carry out the operations and goals of the Company. The predictive power possessed by the
Company tends to adhere to the varied range of laws and policies that have their relevance in
explaining the empirical models by assessing the predictive power. The definition associated
with prediction tends to point towards the empirical prediction rather than prediction carried out
theoretically to achieve the targets and goals of the Company.
Empirical Models for Explanation and Prediction
According to Hudson and Day (2019), the empirical models for explanation and prediction prove
to be beneficial and important to be followed within the organizational sphere by sticking to the
several policies and strategies laid down by predictive analytics. The empirical models
associated with explanation and prediction tends to be included under the purview of one theory
that is established systematically within the organizational sphere. Both these goals and aims are
thought to be desirous and included within the organizational sphere to achieve the objectives
and goals of the organization by aligning with the several practices of predictive analytics.
According to Maxwell et al., (2017), explanation and prediction are considered as two separate
aims that are conceived. A thin line of difference exists between them as the strategies and
operations associated with the explanatory statistical model and the predictor model differ in
their operational actions and services.
The models that are functional within the market tend to devise the specific strategies that have
an impact upon developing the various operations and activities related to the organizations. The
empirical explanation and the analytics associated with the empirical methods differ to establish
the activities and services within the organization.
Analysis
The predictive analytics have a huge role to play and comprises of both the benefits and
drawbacks within their functioning. The efficiency within the production activities is established
with the help of adhering to the practices and steps of predictive analytics. According to Siddiqi
7

(2017), the utilisation of predictive analysis aligned with the predictive power tends to involve
the activities and services that are beneficial and help in recording newer observations within the
market in an accurate manner by specifying the laws and policies.
According to Maxwell et al., (2017), the practices of developing the predictive analytics and its
various concepts takes into account the facts that fraudulent activities are checked and the risks
that are prevalent tend to exist within the organisational sphere that helps in providing the
benefits of the organisation. The policies and the laws are stuck to while implementing the
operational activities properly. Disadvantages also occur within the execution of the operations
associated with the concept of predictive analytics and it tends to affect the collection of data due
to over-fitting the empirical facts tend to fit the training data perfectly to achieve the targets of
the Company in a perfect manner. It will lead to accuracy that is low predictive on the existing
data and facts aligned with the strategies and policies. Having smaller data sets in predictive
analytics can result in less partitioning of the data that would affect the low predictive accuracy
on the data.
8

Conclusion
The roles of predictive analytics and its various features were discussed aligned with the
practices aligned with the specific systems and policies. The roles played by the predictive
analytics integrated with the consolidated research were discussed within the organisational
sphere. The empirical models of explanation and prediction were also established with the
integration of specific strategies and policies being devised by the organisation to achieve the
objectives and targets. The empirical models for explanation and prediction were covered and the
specific details being included to help in establishing the traits of predictive analytics in a
successful manner. The history associated with the practices of predictive analytics was also
covered and the benefits and disadvantages of the activities of predictive analytics were also
covered. The analysis of the empirical models for prediction and the empirical model of
explanation was also covered within the execution of the study.
9

Reference List
Alemi, F., 2018. What Makes Travelers Use Ridehailing?: Exploring the Latent Constructs
Behind the Adoption and Frequency of Use of Ridehailing Services, and Their Impacts on the
Use of Other Travel Modes. University of California, Davis.
Collins, H., 2018. Creative research: the theory and practice of research for the creative
industries. Bloomsbury Publishing.
Harris, S.L., May, J.H. and Vargas, L.G., 2016. Predictive analytics model for healthcare
planning and scheduling. European Journal of Operational Research, 253(1), pp.121-131.
Hudson, V.M. and Day, B.S., 2019. Foreign policy analysis: classic and contemporary theory.
Rowman & Littlefield Publishers.
Kessler, R.C., 2018. The potential of predictive analytics to provide clinical decision support in
depression treatment planning. Current opinion in psychiatry, 31(1), pp.32-39.
Maxwell, S.E., Delaney, H.D. and Kelley, K., 2017. Designing experiments and analyzing data:
A model comparison perspective. Routledge.
McCullagh, P., 2019. Generalized linear models. Routledge.
Neuendorf, K.A., 2016. The content analysis guidebook. Sage.
Parikh, R.B., Kakad, M. and Bates, D.W., 2016. Integrating predictive analytics into high-value
care: the dawn of precision delivery. Jama, 315(7), pp.651-652.
Shah, N.D., Steyerberg, E.W. and Kent, D.M., 2018. Big data and predictive analytics:
recalibrating expectations. Jama, 320(1), pp.27-28.
Shmueli, G. and Koppius, O.R., 2011. Predictive analytics in information systems research. MIS
quarterly, pp.553-572.
Shroff, P., Psychability Inc, 2017. Systems and methods to utilize subscriber history for
predictive analytics and targeting marketing. U.S. Patent Application 15/288,321.
Siddiqi, N., 2017. Intelligent credit scoring: Building and implementing better credit risk
scorecards. John Wiley & Sons.
10