Research Paper

Background

Methods

Results

Discussion and Conclusion

Do not use abbreviations or insert tables, figures or references into your abstract. You abstract generally should not exceed about 300 words.

· This section provides an overview of your project and introduces the background work to it.

· In this section you might wish to include reasons why you feel you are a suitable candidate for performing the project (why you feel you can do it, what skills are required and how you fulfill these requirements), why the topic interests you specifically, and why you chose the project in the first place.

· This section might also include an introduction to the industry or organization being investigated or evaluated.

· Overall, this section will set the scene for the project.

· This section identifies other work, publications and research related to your topic.

· It will demonstrate that your project does not exist in an academic vacuum but relates to other research topics and fields of current interest.

· Related research can also help demonstrate your understanding of your topic area, showing the reader that you are aware of what is currently happening in the field and are conversant with other topics that impinge upon it (Dawson, 2005).

· What is known? Please describe what empirical research tells us about your topic.

· Where is the GAP? Please describe the gap in the field that your study will address. Describe what we need to know that research fails to tell us.

· What will your study contribute to our broader knowledge regarding this topic? (Think about both contributions to scholarship and/or society at large.)

· Your project proposal may also include the research question you intend to investigate and, hopefully, answer to some extent within your project.

· Computing projects do not necessarily set out to answer particular questions,

· But for some projects (particularly research degree projects) a statement of your research question is essential.

Examples of research questions are:

1. Does the size of an organisation affect its commitment to software quality standards?

2. What is the relationship, if any, between software maintainability and coding structure standards?

3. Is there an optimum solution to the prediction of software development costs?

4. How do large organisations maintain quality standards in the development of internal software?

· Aims identify at the highest level what it is you hope to achieve with your project – what you intend to achieve overall.

· An aim is a broad statement of intent that identifies your project’s purpose.

Aim Example:

· Evaluate artificial intelligence techniques for modelling weather patterns.

· Objectives, on the other hand, identify specific, measurable achievements that build towards the ultimate aim of your project.

· They are more precise than aims as they are ‘quantitative and qualitative measures by which completion of the project will be judged’

Objectives Examples:

· Identify and evaluate existing weather pattern modelling techniques.

· Identify artificial intelligence approaches suitable for modelling weather patterns.

· Design and develop at least three artificial intelligent systems for modelling weather patterns.

· Compare and contrast the developed systems with one another and existing approaches to modelling weather patterns.

· This section of your proposal will identify precisely what you intend to submit at the end of the project.

· It may well identify a written report that covers particular points and makes certain recommendations.

· A chapter breakdown may be included where appropriate.

· It can describe programs and user documentation and it might include models and algorithms that will be developed to address specific problems.

· You might also be delivering a functional specification for a piece of software, a prototype, or a test plan.

This project focuses on design and development of an intelligent bank marketing system for a better bank marketing. The project will be carried out in four phases with significant milestone at the end of each phase. In the first phase, we will download and prepare the bank marketing data. Then we will search for information to have clear knowledge about the bank marketing problem and its solutions.

The second phase of the project would concentrate on an analytical framework to analyze the performance of different data mining algorithms that would be developed identifying the main components of the bank marketing.

The third phase would concentrate on the design and the implementation of efficient, reliable and robust data mining tool that achieves a better classification accuracy. Moreover, an ensemble of classifiers would be developed to increase classification performance of an efficient bank marketing.

The fourth and final phase would concentrate on preparing and submitting a conference paper. More details on the different phases are as follows:

First of all, we will download and prepare the bank marketing data for the implementation with WEKA data mining tool. Since the data is unbalanced we try to balance the by using SMOTE technique to have an efficient and accurate classification. Then we will check previous studies done on the same field to have clear knowledge about the bank marketing problem and its solutions and applied methods before.

In this phase, efficient, intelligent and robust bank marketing system will be developed. The system will employ the bank marketing data with robust attribute selection. The extracted attributes will be employed by means of the Data mining techniques for intelligent and more efficient bank marketing environment. This platform will serve to design and develop application oriented bank marketing system. The downloaded data can be processed by employing a variety of feature extraction and attribute selection techniques. Then comparison of different feature extraction, attribute selection techniques, and data mining algorithms for bank marketing will be made. The employed feature extraction and attribute selection techniques should be robust and intelligent, so that a reliable bank marketing can be realized. The proposed platform will be used to understand and implement the intelligent system for a Bank Marketing. The finding of robust features and attributes of bank marketing data would be very meaningful for determining important insights into the effects of various parameters on the performance of bank marketing system to enlighten which feature extraction and attribute selection a is the most effective and less time consuming.

A classification problem is referred to as imbalanced when the instances in one or several classes, known as the majority classes, out number the instances of the other classes, called the minority classes. The synthetic minority over-sampling technique (SMOTE) (Chawla et al., 2002) is a well acknowledged over-sampling method. In the SMOTE, instead of mere data oriented duplicating, the positive classis over-sampled by creating synthetic instances in the feature space formed by the positive instances and their K-nearest neighbours.

In this step, feature extraction, attribute selection and data mining algorithms will be implemented as off-line. The feature extraction and attribute selection methods are in the set of data processing tools which extract features from the bank marketing data. Feature extraction and attribute selection are also one of the most important steps in data classification. It is highly effective technique in selection of attributes and is frequently applied to complex, high dimensional, multivariate data. When the features are not appropriate for the given classification problem, obtained performances will be unsatisfactory. In this case, even the classification algorithm is optimally determined for the problem, because of the improper features/attributes; the algorithm cannot generate high performance. Therefore, it is mandatory to find and extract suitable features from the raw data to be able to obtain good classification results. Many feature extraction and attribute selection techniques will be applied. These are CFS Subset Evaluator, and Infogain Attribute Evaluator etc. will be applied.

The marketing data data is of multi-dimensional nature. Therefore, it is difficult to find a robust feature extraction, attribute selection and data mining algorithm for Bank marketing. Data mining algorithms have ability to distinguish different type of data. In this phase the focus will be to make a comparison of different data mining algorithms in the field of Bank marketing. The outcome will be the analysis and choice of most appropriate sets of features extraction, attribute selection and data mining techniques, for the Bank marketing. After applying different feature extraction/attribute selection algorithms, the existing data mining algorithms (such as ANN, k-NN, SVM, decision tree algorithms, etc.) capable of dealing with network data will be applied, implemented and tested. For this purpose, various classification schemes for a particular data classification task will be developed. The aim of using data mining techniques is to make a better Bank marketing. Even though one of the algorithms would produce the best overall performance, ensemble classifier approach, where the idea is to consider more than one classification scheme can give better classification accuracy. Classifier ensembles are multiple classifier systems trained on different data or feature subsets, will be used to get better performance and accuracy.

The fourth and final phase would concentrate on preparing and submitting a conference paper. Here, the performance metrics such as Area under ROC curve, F-measure, kappa statistic and total classification accuracy would be quantified with the help of WEKA software. Obtained results from experimentation would then be used to verify the accuracy, reliability and robustness of the developed models and would provide feedback for improvement and fine-tuning the models in phase 1, phase 2 and phase 3. The obtained results of the ensemble classifiers would be benchmarked with classical single classifiers. Moreover efficient, intelligent and robust methods will be achieved for bank marketing.

Use headings 2 and 3 as appropriate, and use these headings if appropriate.

· Use case Diagrams

· Database Diagrams

· Relation Diagrams

· describes the research and project methods you will use in performing your project.

· should not identify methods that you might be investigating as part of your project, but those methods you are actually using.

· Include development methods that you are using as part of a systems development; survey methods for a case study evaluation and evaluation methods for comparing two or more systems.

· Research methods include action research, case study, survey and experiment.

· Give the tables and results of your research project one by one

· Give the explanation on the results.

· Give discussion on the results of your research project

· Give comments on the results.

· Conclude and summarize the results of your research project.

Proposed steps schedule is planned like in the table.

Activity-on-the-node diagram represents the tasks you are performing in your project as nodes connected by arrows (Dawson, 2005)

· You might need to identify any resource requirements for your project, such as hardware, software and access to particular computers.

· If you have access to particular resources, this fact should be pointed out in this section.

· If the resources for your project are not available in your department, or are too expensive to obtain, your project will be unacceptable.

· However, if you know you need a particular piece of software or hardware, you must find out its cost and include this information within this section.

· A proposal that omits this information may be rejected if the assessor does not know how inexpensive or available the item is and might assume it is beyond your project’s budget.

Dawson, C. W. (2005). Projects in computing and information systems: a student’s guide. Pearson Education.

Use the APA Style of referencing.

Check the related slides and Rubric for the project report details