Business Analytics Assignment: Predictive Analytics in Health Care

Question

Task: Business Analytics Assignment Instructions:
Perform a literature review on a known topic in business analytics. It can be any topic on tools, methodologies or applications.

Some examples include, but not limited to:
1. Use of predictive analysis in healthcare industry
2. Comparison of BI tools
3. Techniques of predictive analysis
4. Methods of representing multi-dimensional data in visualizations
5. Analytics techniques to improve logistics management
6. Security of data and privacy concerns in analytics

Answer

Introduction to the context of business analytics assignment:
When we have a large collection of data that we would like to do statistical testing or form acknowledgment, “machine learning” (ML) is the path ahead.ML is the quickest-growing field in “computer science” (CS), and is a big challenge. CS The goal of ML is to create algorithms that can learn and improve around time and be able to be used for projections. ML methods are commonly used in a number of areas, and the health care sector in particular has benefited immensely from ML prediction techniques. This provides a variety of notifying and risk assessment choice-making methods aimed at maximizing patient protection and health outcomes. With the necessity to minimize health care expenses and drive in the direction of customized health care, the health care business confronts obstacles in key fields such as electronic record keeping, digital convergence, and computer-aided diagnostics and disease prediction.ML includes a broad variety of methods, strategies and systems to overcome these problems.

Literature Review:
Industrial companies are focused on large volumes of data that need to be understood through Machine Learning. Through gaining information from these results, companies are able to function more effectively and gain leverage over their rivals.Unconventional predictive models has been widely extended to many areas of ML algorithms. ML methods and technologies are used in tasks such as “browsing, advertisement, YouTube. Health Car Informatics, a multi-disciplinary field, has become synonymous with technical advances and difficulties in data management” [1]. Health Informatics is a technological restraint that contracts with the collection, retrieval and efficient use of medical records, data and expertise for problem-solving and decision-making. Health Technology has grown enormously over the years, such as developments in knowledge processing, diagnosis, communications and analysis.PA is a category of PA which is used to make decisions about unpredictable events in the future. PA uses a range of methods from data analysis, statistics, simulation, computer learning and artificial intelligence to analyze recent results in order to make forecasts about the future.The forecaster is the key PA agent that is known as the component used to calculate predicted actions” [2]. With the aid of predictors, potential outcomes are predicted with extremely accurate results. Methods used to perform analytical modeling can be categorized as ML and RT. ML approaches have become progressively more popular in the conduct of PL due to their excellent accomplishment in managing huge-level datasets with complex traits and boisterous data.Such simulations are commonly used in predictive data processing applications such as market estimation, risk management, predictive consumer behavior, and content classification” [3].

Predictive Analytics supports various divisions of wellbeing “life sciences” and suppliers. It helps to detect diseases reliably, enhance patient safety, maximize services and also increase health outcomes.Predictive Modeling lets companies’ budget for healthcare by reducing costs [4]. The development of predictive analytics in this sector is expected to produce a positive result by enhancing service efficiency. “Predictive Analytics” (PA) has the potential of changing the healthcare market.

According to Cirkovic [5] ML: the typical concept is-A CP is told to discover from knowledge in relation to any category of duties and output assessment , if its success in tasks T, as measured by P, increases with knowledge ”. ML is an contrived intellect division that utilizes a range of mathematical, and optimization methods allowing pcs to understand from previous instances and identify patterns that are difficult to distinguish from large, chaotic, or complicated information sets. ML is a data modeling system that automates computational model construction.Machine learning helps computers to discover secret insights all the way through processes that iteratively understand from results, without specifically programming where to look.Enterprises are inspired by the large quantities of data they produce and recycle every day to achieve deeper significance. Efficient algorithms, implementations, and systems are developed for machine learning to gain greater statistical precision and reliability for the data sets of organizations and lead to productive distinct approaches. ML methodologies are developed to recognize the probability of making choices centered on the analytical potential of broad data sets. It shows how efficient it is in performing predictive tasks like determining which patterns have the greatest propensity to produce desired outcomes [6].

Measures to utilize ML to data
ML task can be broken down into under these stages. Data collection: if the information is printed on broadsheet, registered Data must be compiled in an electronic format suitable for review in text files and spreadsheets or stored in a SQL database. This data should function as a learning material and be used by an algorithm to produce actionable information. Data discovery and processing: The success of any machine learning program depends primarily on the quality of the data it uses. This stage of machine learning seems to involve a lot of human interaction. Well-quoted statistics show that 80 per cent of machine learning research is devoted to success.

Data model training: the simple machine learning function must notify the appropriate algorithm selection and the algorithms should reflect the information well into the context of a document.

Evaluating the consistency of the software: Determine how much the algorithm has learned from its previous experience is very critical, because each machine learning software usually results in a skewed solution to the learning problem. The accuracy of the model will be tested according to the type of model used in the comparison collection.

Enhancing model efficiency: it is essential to use sophisticated approaches to improve model efficiency if better performance is desired. It may be appropriate to shift to a specific form of model as a whole every now and then.

Fig.1. ML Process [7]

According toDrukker, Caroline [8] If these measures have been taken, if the model seems to be operating in an appropriate way, it may be implemented for its planned purpose. The brand may be used to offer data for forecasts, financial data estimates, to produce relevant observations for marketing or analysis, or to automate activities.

A predictive model that can be utilized for assignments that include calculating single value utilizing further principles in datasets. When detailed feedback is given to predictive models as to what they need to know and how they are supposed to learn, the predictive model training process is known as supervised learning [9].The ultimate goal of closely controlled ML is “to construct a prototype centered on proof of the presence of ambiguity to make predictions. Precisely, a controlled knowledge process uses a established of key data and recognized data response (productivity) and creates a standard for making accurate predictions for responding to new data”.

Fig. 3. SL [10]

ML is a rapidly increasing technique in the healthcare field.This ML system will also help health professionals evaluate data to find consequences that could lead to better diagnosis and care. IBM Work Group ML for HC and LS is developing and utilizing ML and “data mining” techniques to a complex range of challenges from “clinical genomic analysis” to the implementation of clinical decision support systems [11].

Conclusion:
ML has been turned over the past two decades from academic work into a multi-billion-dollar business and a focal goal for our political, industrial, technological and safety structure. Plentiful research in ML has ignited concern in science of optimization, inspired by large-scale application involving the study of massive, high-dimensional data. ML has provided care practitioners with new resources for communicating with modern methods of treating medicine. This also indicates that machine learning methods and strategies are crucial in the field of wellbeing and are used primarily in the detection and analysis of different kinds of cancers.

Reference
[1] Al-Hagery, M. A. (2015). Knowledge discovery in the data sets of hepatitis disease for diagnosis and prediction to support and serve community. Int. J. Comput. Electron. Res, 4(6), 118-125.

[2] Akutekwe, A., Seker, H., &Iliya, S. (2014). An optimized hybrid dynamic Bayesian network approach using differential evolution algorithm for the diagnosis of Hepatocellular Carcinoma. Paper presented at the 2014 IEEE 6th International Conference on Adaptive Science & Technology (ICAST).

[3] Anand, A., & Shakti, D. (2015). Prediction of diabetes based on personal lifestyle indicators. Paper presented at the 2015 1st International Conference on Next Generation Computing Technologies (NGCT).

[4] Bahl, R., Spolia, S., & Sharma, C. M. (2015). Predicting recurrence in cervical cancer patients using clinical feature analysis.Business analytics assignment Journal of Advances in Medicine and Medical Research, 908-917.

[5] Cirkovic, B. R. A., Cvetkovic, A. M., Ninkovic, S. M., &Filipovic, N. D. (2015). Prediction models for estimation of survival rate and relapse for breast cancer patients. Paper presented at the 2015 IEEE 15th International Conference on Bioinformatics and Bioengineering (BIBE).

[6] Das, J., Gayvert, K. M., & Yu, H. (2014). Predicting cancer prognosis using functional genomics data sets.Business analytics assignment Cancer informatics, 13, CIN. S14064.

[7] Farrar, C. R., & Worden, K. (2012). Structural health monitoring: a machine learning perspective: John Wiley & Sons.

[8] Drukker, Caroline A, "Optimized outcome prediction in breast cancer by combining the 70-gene signature with clinical risk prediction algorithms", Breast cancer research and treatment 145.3(2014): 697-705.

[9] Cancer Informatics 2014:13(S2) 19–28 doi: 10.4137/CIN.S13788.

[10] Charles R. Farrar, Keith Worden, ?Structural Health Monitoring, A Machine Learning Perspective?, John Wiley & Sons, Ltd., Publication, ISBN: 978-1-119-99433-6.

[11] Chi-Chang Chang, Sun-Long Cheng, Chi-Jie Lu and Kuo-Hsiung Liao, ?Prediction of Recurrence in Patients with Cervical Cancer Using MARS and Classification?, International Journal of Machine Learning and Computing, Vol. 3, No. 1, February 2013.

[12] Drukker, Caroline A, "Optimized outcome prediction in breast cancer by combining the 70-gene signature with clinical risk prediction algorithms", Breast cancer research and treatment 145.3(2014): 697-705.