Category: Machine Learning

literature review is *AI/ML-based Auto Graders for Unstructured Open-ended Questions*. The criteria for evaluation and submission requirements are detailed below.
A literature review is a comprehensive summary of previous research on a topic. It involves critical analysis, categorization, and synthesis of relevant scholarly materials and should not merely summarize the sources, but also organize and present them in a way that builds an argument or case for a new research opportunity.
A literature review selects relevant past literature and Connects, Synthesizes, and Evaluates the 25 to 50 most recently published papers within 2023-2024 timeframe, approaches, or articles that are similar or related, putting the authors in conversation with each other in 4000 or more words. You need to identify and download 25-50 recent peer-reviewed articles published in 2023-2024 about your topic to connect, synthesize, and evaluate the information.
An effective literature review fulfills three purposes:
1) Situates your work within the broader scholarly community – connects your work to the broader field and shows that your work has grown organically from current trends.
2) Shows that you have “done your homework” (and builds your credibility) by illustrating your familiarity with the major agreements, debates, and critical findings of the field.
3) Illustrates a “gap” in this previous research—a gap that can be filled by your unique and novel research contribution.
Most of the literature reviews you will encounter are introductory sections to research papers (i.e., proposals, journal articles, theses, etc.), but you may also encounter stand-alone literature reviews that summarize the state of research in a particular field or topic area. It is important to understand the purpose of your literature review so check with a professor if you have questions.
How Do I Write a Literature review?
An effective literature review CONNECTS and GROUPS relevant research based on common themes or trends. Each paragraph should discuss one specific trend, not one specific author.
In general, the topic sentences in a literature review should illustrate the connection across multiple studies: the common agreement/disagreement, the similar focus or the related limitation. Below are some common phrases for connecting studies by showing agreement and disagreement
Use the following phrases to highlight agreement.
o “One trend in the research is….”
o “Research seems to agree that….”
o “Numerous authors support the claim that….”
o “There is strong convergent evidence for….”
Or disagreement
o “The evidence on X is mixed for….”
o “There is overall debate regarding….”
o “A lack of consensus exists on the point of…”
o “There are two conflicting camps on the issue of….”
Below is a brief annotated example of how to write a literature review. However, to provide a meaningful literature review and debate on your assigned topic, you need to include more detailed discussions, resources, and analyses. Your review should be 4,000 words or more and include diagrams, tables, equations, and references. Additionally, you must adhere to the required formatting guidelines and the structured layout (including sections and subsections) of your literature review.
Due to the sensitive nature of passwords, acquiring high-quality password corpora for analysis can be difficult [Summary of research trends]. Data sets that were used in previous work on passwords have all been non-ideal in at least one dimension.
A review of the literature shows that many researchers use password corpora collected from various security leaks [14, 24, 26, 35, 53, 55] [Research trend]. [Assesses limitations of previous research] These corpora tend to be very large (tens of thousands to millions), and they represent in-use passwords selected by users. While this approach has many benefits, these passwords come with no contextual information about how they were made or used, and the released lists are difficult to verify. Furthermore, the largest leaks thus far have come from low-value accounts with weak password-composition policies, such as the RockYou gaming website. In addition, if the password file is encrypted or hashed, only those…
On the other hand, there is an extensive literature that describes researchers asking users to self-report password information [Research trend], including both password composition details and user sentiment information, instead of collecting passwords expressly for an experiment [7,33,37,47, 49,58]. [Assesses limitations of all previous research] While self-reported data can be very useful and can provide a lot of context, it cannot always be considered reliable, particularly with regard to a sensitive topic like passwords.
Finally, a small number of researchers have been able to work with large organizations to collect authentic data [Research trend]. [Indicates method weakness with all studies] Florêncio and Herley used an opt-in component of the Windows Live toolbar to collect … Bonneau worked with Yahoo! to analyze plaintext passwords … Both studies include very large, reliable samples, as well as good contextual information. Due to security concerns, however, in both studies researchers were able to record only extremely limited information about the content of the passwords, precluding many…. [Connects to current study] In contrast, our paper connects information about each user with analysis of that user’s plaintext….
In this paper, we overcome many limitations of past studies. [Illustrates how current study overcomes limitations of previous research] Our password corpus includes more than 25,000 real passwords, created by users for frequently used, high-value accounts unrelated to our research context. We have indirect, yet extensive, access to plain-text passwords, allowing us to perform more complex and thorough analyses than was possible for other similarly authentic corpora.
Required Formatting
1) Resources: You need to identify and download 25-50 recent peer-reviewed articles published in 2023-2024 about your topic to connect, synthesize, and evaluate the information.
2) Pages/Words: There’s no one-size-fits-all answer. The length should be dictated by the breadth of the topic and the requirements of the assignment or publication. Typically, literature reviews are in 4000 or more words but can be much longer in some cases. Please note that submitting less than 4000 words will result in a grade of zero for this assignment.
3) Margins: Use 1-inch margins on all sides.
4) Line Spacing: Double-spacing is standard for the body text; single-spacing can be used for block quotations and references.
5) Paragraphs: Use a first-line indent for paragraphs; block format can be used for headings and subheadings.
6) Font Type/Family: Use a standard, easily readable font. Times New Roman, Arial, and Calibri are common choices in academic writing.
7) Font Size: 12pt for the main text is standard; 10-11pt can be used for footnotes and figure captions.
8) Number of Figures: Use figures and tables only as needed to illustrate key points or summarize information.
9) Captions: Every figure and table should have a caption that describes its content and relevance to the literature review.
10) Citation Style: Use a consistent citation style throughout your review. APA, MLA, and Chicago are common.
11) Reference List: All sources cited in the text should appear in the reference list at the end of your document.
12) Proofreading: Check for spelling and grammar errors and ensure clarity and coherence in your writing.
13) Plagiarism: Make sure to paraphrase correctly and cite all sources to avoid plagiarism.
Required Structure (section titles, but feel free to add more subsections)
1) Topic Title: Title, your name, and your institution/college/department.
2) Abstract: literature reviews start with an abstract that summarizes the key points.
3) Introduction: Define the topic, provide an overview of the review, and state the objectives or research questions.
4) Body: Organize the literature into subsections that present themes or identify trends, including relevant theories, methods, and results.
a. Relevant Theories
b. Methods
c. Results
5) Conclusion: Summarize the major contributions, note the current state of research, and suggest areas for future work.
6) References: Use a consistent citation style throughout your review. APA, MLA, and Chicago are common. Your resources must include studies published recently, specifically within the 2023-2024 timeframe.
Remember that a literature review is not just a list of summaries, but a critical analysis that makes sense of the corpus of literature. It should identify gaps, inconsistencies, and the need for additional research. The formatting should support the readability and organizational structure of the document. Always adhere to the listed above guidelines provided.

literature review is *AI/ML-based Auto Graders for Unstructured Open-ended Questions*. The criteria for evaluation and submission requirements are detailed below.
A literature review is a comprehensive summary of previous research on a topic. It involves critical analysis, categorization, and synthesis of relevant scholarly materials and should not merely summarize the sources, but also organize and present them in a way that builds an argument or case for a new research opportunity.
A literature review selects relevant past literature and Connects, Synthesizes, and Evaluates the 25 to 50 most recently published papers within 2023-2024 timeframe, approaches, or articles that are similar or related, putting the authors in conversation with each other in 4000 or more words. You need to identify and download 25-50 recent peer-reviewed articles published in 2023-2024 about your topic to connect, synthesize, and evaluate the information.
An effective literature review fulfills three purposes:
1) Situates your work within the broader scholarly community – connects your work to the broader field and shows that your work has grown organically from current trends.
2) Shows that you have “done your homework” (and builds your credibility) by illustrating your familiarity with the major agreements, debates, and critical findings of the field.
3) Illustrates a “gap” in this previous research—a gap that can be filled by your unique and novel research contribution.
Most of the literature reviews you will encounter are introductory sections to research papers (i.e., proposals, journal articles, theses, etc.), but you may also encounter stand-alone literature reviews that summarize the state of research in a particular field or topic area. It is important to understand the purpose of your literature review so check with a professor if you have questions.
How Do I Write a Literature review?
An effective literature review CONNECTS and GROUPS relevant research based on common themes or trends. Each paragraph should discuss one specific trend, not one specific author.
In general, the topic sentences in a literature review should illustrate the connection across multiple studies: the common agreement/disagreement, the similar focus or the related limitation. Below are some common phrases for connecting studies by showing agreement and disagreement
Use the following phrases to highlight agreement.
o “One trend in the research is….”
o “Research seems to agree that….”
o “Numerous authors support the claim that….”
o “There is strong convergent evidence for….”
Or disagreement
o “The evidence on X is mixed for….”
o “There is overall debate regarding….”
o “A lack of consensus exists on the point of…”
o “There are two conflicting camps on the issue of….”
Below is a brief annotated example of how to write a literature review. However, to provide a meaningful literature review and debate on your assigned topic, you need to include more detailed discussions, resources, and analyses. Your review should be 4,000 words or more and include diagrams, tables, equations, and references. Additionally, you must adhere to the required formatting guidelines and the structured layout (including sections and subsections) of your literature review.
Due to the sensitive nature of passwords, acquiring high-quality password corpora for analysis can be difficult [Summary of research trends]. Data sets that were used in previous work on passwords have all been non-ideal in at least one dimension.
A review of the literature shows that many researchers use password corpora collected from various security leaks [14, 24, 26, 35, 53, 55] [Research trend]. [Assesses limitations of previous research] These corpora tend to be very large (tens of thousands to millions), and they represent in-use passwords selected by users. While this approach has many benefits, these passwords come with no contextual information about how they were made or used, and the released lists are difficult to verify. Furthermore, the largest leaks thus far have come from low-value accounts with weak password-composition policies, such as the RockYou gaming website. In addition, if the password file is encrypted or hashed, only those…
On the other hand, there is an extensive literature that describes researchers asking users to self-report password information [Research trend], including both password composition details and user sentiment information, instead of collecting passwords expressly for an experiment [7,33,37,47, 49,58]. [Assesses limitations of all previous research] While self-reported data can be very useful and can provide a lot of context, it cannot always be considered reliable, particularly with regard to a sensitive topic like passwords.
Finally, a small number of researchers have been able to work with large organizations to collect authentic data [Research trend]. [Indicates method weakness with all studies] Florêncio and Herley used an opt-in component of the Windows Live toolbar to collect … Bonneau worked with Yahoo! to analyze plaintext passwords … Both studies include very large, reliable samples, as well as good contextual information. Due to security concerns, however, in both studies researchers were able to record only extremely limited information about the content of the passwords, precluding many…. [Connects to current study] In contrast, our paper connects information about each user with analysis of that user’s plaintext….
In this paper, we overcome many limitations of past studies. [Illustrates how current study overcomes limitations of previous research] Our password corpus includes more than 25,000 real passwords, created by users for frequently used, high-value accounts unrelated to our research context. We have indirect, yet extensive, access to plain-text passwords, allowing us to perform more complex and thorough analyses than was possible for other similarly authentic corpora.
Required Formatting
1) Resources: You need to identify and download 25-50 recent peer-reviewed articles published in 2023-2024 about your topic to connect, synthesize, and evaluate the information.
2) Pages/Words: There’s no one-size-fits-all answer. The length should be dictated by the breadth of the topic and the requirements of the assignment or publication. Typically, literature reviews are in 4000 or more words but can be much longer in some cases. Please note that submitting less than 4000 words will result in a grade of zero for this assignment.
3) Margins: Use 1-inch margins on all sides.
4) Line Spacing: Double-spacing is standard for the body text; single-spacing can be used for block quotations and references.
5) Paragraphs: Use a first-line indent for paragraphs; block format can be used for headings and subheadings.
6) Font Type/Family: Use a standard, easily readable font. Times New Roman, Arial, and Calibri are common choices in academic writing.
7) Font Size: 12pt for the main text is standard; 10-11pt can be used for footnotes and figure captions.
8) Number of Figures: Use figures and tables only as needed to illustrate key points or summarize information.
9) Captions: Every figure and table should have a caption that describes its content and relevance to the literature review.
10) Citation Style: Use a consistent citation style throughout your review. APA, MLA, and Chicago are common.
11) Reference List: All sources cited in the text should appear in the reference list at the end of your document.
12) Proofreading: Check for spelling and grammar errors and ensure clarity and coherence in your writing.
13) Plagiarism: Make sure to paraphrase correctly and cite all sources to avoid plagiarism.
Required Structure (section titles, but feel free to add more subsections)
1) Topic Title: Title, your name, and your institution/college/department.
2) Abstract: literature reviews start with an abstract that summarizes the key points.
3) Introduction: Define the topic, provide an overview of the review, and state the objectives or research questions.
4) Body: Organize the literature into subsections that present themes or identify trends, including relevant theories, methods, and results.
a. Relevant Theories
b. Methods
c. Results
5) Conclusion: Summarize the major contributions, note the current state of research, and suggest areas for future work.
6) References: Use a consistent citation style throughout your review. APA, MLA, and Chicago are common. Your resources must include studies published recently, specifically within the 2023-2024 timeframe.
Remember that a literature review is not just a list of summaries, but a critical analysis that makes sense of the corpus of literature. It should identify gaps, inconsistencies, and the need for additional research. The formatting should support the readability and organizational structure of the document. Always adhere to the listed above guidelines provided.

I have all answers i need edit and change some valus
check plagiarism andAi
Note: in attched assignment questions and requirements
i will upload assignment answers later to edit on it

The presentation is Genomic Data in Disease Prediction and Prevention. The criteria for evaluation and submission requirements are detailed below.
In 60 or more slides, prepare your PowerPoint slides and presentation about your assigned topic, which is Genomic Data in Disease Prediction and Prevention. Please note that submitting less than 60 slides will result in a grade of zero for this optional assignment. As a general rule, try to keep presentations as concise as possible while covering all necessary information. For example, a 60-minute presentation might effectively utilize between 60 and 90 slides, depending on the factors mentioned below.
Required slide titles and contents.
Use elements such as questions, short videos, or interactive objects to engage the audience and make the presentation more interactive. Also, please note that each of the slide titles listed below can be presented across one or more slides.
1. Title Slide
– Include a clear, concise title that immediately gives an idea of the subject matter.
– Add your full name, your institution/college/department, and the date of the presentation.
2. Introduction to Genomic Data in Disease Prediction and Prevention
– Provide 1-3 succinct definitions to clarify what Genomic Data in Disease Prediction and Preventionis.
– Discuss the historical development of Genomic Data in Disease Prediction and Prevention, including information about its invention and key milestones.
3. Classification of Machine Learning Relevant to Genomic Data in Disease Prediction and Prevention
– Outline which categories of machine learning (supervised, unsupervised, reinforcement) Genomic Data in Disease Prediction and Preventionfalls into and why.
4. Architectural Overview of Genomic Data in Disease Prediction and Prevention
– Present a diagram showing the architecture of Genomic Data in Disease Prediction and Prevention, explaining the components and how they interact.
5. Operational Mechanics of Genomic Data in Disease Prediction and Prevention
– Describe the process of how Genomic Data in Disease Prediction and Preventionworks, using a step-by-step approach.
– Include diagrams or flowcharts to visually represent each step.
6. Parameters and Hyperparameters of Genomic Data in Disease Prediction and Prevention
– List the main parameters and hyperparameters that influence the operation of Genomic Data in Disease Prediction and Prevention.
– Explain the impact of each parameter on the performance of Genomic Data in Disease Prediction and Prevention.
7. Python Code Snippets for Configuring Genomic Data in Disease Prediction and Prevention
– Provide actual code examples showing how to set up and configure the parameters and hyperparameters of Genomic Data in Disease Prediction and Prevention.
8. Derivatives Associated with Genomic Data in Disease Prediction and Prevention
– Explain any mathematical derivatives used in the functioning or optimization of Genomic Data in Disease Prediction and Prevention.
9. Mathematical Framework of Genomic Data in Disease Prediction and Prevention
– Present key equations and formulas used in Genomic Data in Disease Prediction and Prevention, explaining each component’s role and significance.
10. Heuristic Approaches to Modeling Genomic Data in Disease Prediction and Prevention
– Discuss heuristic models used in Genomic Data in Disease Prediction and Prevention, explaining their benefits and limitations.
11. Data Handling Capabilities of Genomic Data in Disease Prediction and Prevention
– Detail the types of data Genomic Data in Disease Prediction and Preventioncan process (e.g., numerical, categorical, images).
– Mention the typical data sizes Genomic Data in Disease Prediction and Preventionis capable of handling efficiently.
12. Analyzing the Strengths and Weaknesses of Genomic Data in Disease Prediction and Prevention
– List and explain the advantages and limitations of using Genomic Data in Disease Prediction and Preventionin various scenarios.
13. Identifying and Addressing Overfitting and Underfitting in Genomic Data in Disease Prediction and Prevention
– Describe what signs indicate overfitting or underfitting in Genomic Data in Disease Prediction and Prevention.
– Suggest methods or techniques to address these issues.
14. Practical Applications of Genomic Data in Disease Prediction and Prevention
– Showcase real-world applications of Genomic Data in Disease Prediction and Prevention, illustrating its relevance and utility.
15. Challenges Encountered with Genomic Data in Disease Prediction and Prevention
– Discuss common challenges and obstacles faced when implementing or operating Genomic Data in Disease Prediction and Prevention.
16. Evaluation Metrics for Assessing Genomic Data in Disease Prediction and Prevention
– List the metrics used to evaluate the effectiveness of Genomic Data in Disease Prediction and Prevention.
– Provide a brief explanation of each metric.
17. Cost Function Analysis for Genomic Data in Disease Prediction and Prevention
– Describe the cost functions commonly used with Genomic Data in Disease Prediction and Prevention, highlighting how they influence outcomes.
18. Optimization Algorithms for Genomic Data in Disease Prediction and Prevention
– Detail the optimization algorithms that work best with Genomic Data in Disease Prediction and Prevention, discussing their benefits.
19. Fine-Tuning Strategies for Genomic Data in Disease Prediction and Prevention
– Offer strategies for fine-tuning Genomic Data in Disease Prediction and Preventionto enhance performance, including practical tips or adjustments.
20. Essential Python Libraries for Genomic Data in Disease Prediction and Prevention
– List key Python libraries used with Genomic Data in Disease Prediction and Prevention.
– Provide brief installation and usage instructions for each library.
21. Learning Genomic Data in Disease Prediction and Prevention: Recommended Tutorials
– Include links to friendly tutorials and resources that offer a good introduction to Genomic Data in Disease Prediction and Prevention.
22. Advanced Resources for Genomic Data in Disease Prediction and Prevention
– Recommend top articles, textbooks, and online materials for deeper insights into Genomic Data in Disease Prediction and Prevention.
23. Accessing Full Python Source Code for Genomic Data in Disease Prediction and PreventionImplementations
– Provide a link or QR code that attendees can use to access the full source code of Genomic Data in Disease Prediction and Preventionimplementations.
24. Open Research Questions in Genomic Data in Disease Prediction and Prevention
– Highlight current research gaps or unanswered questions in the field of Genomic Data in Disease Prediction and Prevention.
25. Summary and Conclusion
– Recap the key points covered in the presentation.
– Summarize the potential future developments and outlook for Genomic Data in Disease Prediction and Prevention.
26. Q&A Session
– Invite questions from the audience to clarify topics covered or explore related areas further.
27. References
– Cite all sources used in the preparation of the presentation in an appropriate format.
Required slide formatting.
A good PowerPoint slideshow complements your presentation by highlighting your key message, providing structure, and illustrating important details. While it is not difficult to create a good PowerPoint presentation, it is very easy to create a bad one. Bad PowerPoint presentations may have one or more of the following characteristics: too much specialized detail, too many slides, too many colors, unnecessary images or effects, small text, unreadable figures, and/or unclear slide order.
The strategies below can help you to create effective presentations and to save your audience from “death by PowerPoint.”
Creating Slides
The classic PowerPoint error is to write sentences on a slide and read them. Rather than treating your slides as a script for your presentation, let the content on your slides support your message. Remember: LESS IS MORE.
Keep It Simple and Clear
Text
ØWhere possible, include a heading for each slide.
ØAim for no more than 6-8 lines of text per slide.
ØLimit bullet points to 4-6 per block of text and avoid long sentences.
ØFont size: 30 – 48 point for titles, 24 – 28 for text
ØAvoid all capital letters.
ØMaintain the same font size and style for all slide headers.
ØUse a small font size (e.g., 12pt) at the bottom of the slide for references.
ØProofread carefully for spelling and grammar.
Figures and Images
ØEnsure images are clear and relevant.
ØLabel all figures and tables.
ØPut units beside numbers on graphs and charts.
General Design Principles
ØEmbrace empty space.
ØUse vertical and horizontal guide markers to consistently align elements.
ØAvoid too many colors, clutter or fancy visual effects.
ØUse high contrast to ensure visibility: e.g. Black text on white background or black on light blue.
ØMaintain consistency of the same elements on a slide (colors, fonts, styles, placement etc.), as well as, between slides in the slide deck
ØUse animation sparingly, if at all. If you use transitions, use the same kind each time.
ØEdit entire slide deck to ensure organization is logical and design is consistent.
ØUse a small font size (e.g., 12pt) at the bottom of the slide for references.
ØMaintain consistency in the citation style throughout the presentation.
ØDouble-check for typos and grammatical errors.

My idea in my research is to develop a browser extension that examines the entire content of emails to detect potential phishing emails and alert users in real-time about suspicious or malicious emails, then evaluate the efficacy of the extension using machine learning model metrics such as True Positive, True Negative, Accuracy, and Rate Score to assess its effectiveness in optimizing phishing detection features and ensuring user-friendliness.
So I would like to ask for your assistance in preparing the ‘backgrounds’ section for the research paper “IEEE” as the attachment file.
“backgrounds” can be used to refer to the fundamental and historical information that the reader needs to understand before delving into the reading or study of previous research (related work). This information may include basic concepts, an overview of the field, the history of research in the field, and any significant developments that have occurred.
I need from you to add this things in the background:
+details about machine learning techniques, including deep learning and other types.
Explain also the process like select the model then dataset and trainings and testing methodology that people use in machine learning +details about email and what the vulnerabilities of emails and how analysis the header such as header has important information and what the vulnerability.
Also use academic word,and but reverence about research papers that use it.

My problem statement is In the face of ever-evolving malware, cybersecurity demands robust solutions that can effectively detect and predict malicious activities. While traditional methods have limitations, deep learning offers promising advancements. However, there is a need for a comparative study on the efficacy of various deep learning architectures, including Multi-Layer Perceptron (MLP) classifiers, in conjunction with powerful frameworks like TensorFlow and Keras, for malware detection and prediction. The study aims to implement and evaluate the performance of these models against benchmarks, focusing on the ability to handle complex patterns and adapt to changing malware characteristics. Utilizing a combination of static and dynamic analysis, the project will explore feature extraction and representation learning to optimize model performance. The ultimate goal is to develop a versatile and scalable system that can dynamically update its detection capabilities to counteract emerging malware threats.
I need complete project while involves providing dataset files,csv files, methodology, dataset preparation,preprocessing, features extraction,ppts, discussions, and conclusions. Provide programming code for the problem statement
I will provide sample files and links please follow it and complete the project

Implement a basic neural network using TensorFlow/Keras to classify images from the MNIST dataset. The network should have one hidden layer with 128 neurons and use ReLU activation function. Use softmax activation function for the output layer. Train the network for 10 epochs with a batch size of 32 and evaluate its performance on the test set.

1. Your Python code: def BaumWelch(N,A,B,Pi,T) :

return [A,B,Pi] # Trained Model
Where:
N: Number of model states A: Transition probability matrix
B: Output probability matrix Pi: Initial state probability matrix
T: Training data Return re-estimated A, B and Pi matrices
Notes:
1. Your code must be original work – That is: it must not be a copy of code found online.
2. Submit your python code in a separate file named “viterbi_.py”
3. As Outlook filters out executable files, submit your work in a .zip file
2. A 2 page write-up describing the Baum-Welch algorithm and your implementation of it
Note: Be sure to list all references
3. Your hand calculations for the first observation vector
4. A sample run for each of the three observation vectors
Note: Your sample run for the 1st two observation vectors must match your hand calculations.
Please Note:
1. Your code must be original work – That is: it must not be a copy of code found online.
2. All references must be cited.
3. No scans of hand-written calculations for any part of the assignment will be accepted.
4.The sample runs may be screen-grabbed.

INSTRUCTIONS
Every learner should submit his/her own homework solutions. However, you are allowed to discuss the homework with each other– but everyone must submit his/her own solution; you may not copy someone else’s solution.
The homework helps you understand and apply K Nearest Neighbor.
Follow the prompts in the attached Jupyter notebook. CAP4611-HW3-NearestNeighbors.ipynbDownload CAP4611-HW3-NearestNeighbors.ipynb
Download the data from (Modules/ Data for Homework Assignment) and place it in your working directory, or modify the path to upload it to your notebook.
Before every code cell, add markdown cells to your analysis. Include your solutions, comments, and answers on how to solve the problem. Add as many cells as you need, for easy readability comments when possible.
Hopefully this homework will help you develop skills, make you understand how K nearest neighbor works.
Submission: Save your ipynb file named with your Name_HW3 (e.g.John_Doe_HW3.ipynb).
Good luck!

Here is my Research Topic:
Predicting Bank Failures in USA
Here is my Research Question:
Which U.S. state is most likely to experience the greatest number of bank failures in the future, and what are the main factors that contribute to this likelihood?
Here is my Dataset
Dataset Name: FDIC Failed Bank List
FDIC Failed Bank List Dataset
Download the FDIC Failed Bank List (CSV)
Here is my chosen ML Method:
Classification machine learning method
Revise and Evaluate Data Analysis Model
In this milestone, you will perform an evaluation of your data analytic model and revise your decision model as needed. For the revision, you can add in additional machine learning models or do feature engineering. You can also create confusion matrices and check for accuracy, precision, recall, or F-Measures. You can do sensitivity analyses, create ROC curves, check error rates and variable selection/feature selection. Please do see the image below for some other options for revisions:
Deliverable
For milestone 4, please ensure you have the following REQUIRED sections ONLY:
Final Research Question: Please state your final research question and describe how it evolved, if it changed, from your Milestone 1 version.
Your final research question should reflect the actual analysis you conducted and the actual insight or prediction you made. For example, if you actually ended up doing a classification task to predict the value of a response variable based on some other predictor variables, that’s what you should state as your precise, quantitative research question.
Model Revision: You should discuss how you revised your model(s) and perhaps how you narrowed down the scope of the project to something coherent and managable. In addition, you should give the details of your final machine learning model and present your final results.
Concomitantly, you should assess the robustness of your model, especially noting what makes it strong or what breaks it or the distribution of uncertainty in it. As such, you’ll likely want to utilize some variant of sensitivity analysis, depending on your particular model, to demonstrate true understanding of what the model’s doing rather than just a rote implementation thereof.
Final Results and Initial Version of Final Conclusion: You should show the initial validation of metrics and also decide upon any additional machine learning algorithms you might want to try. In addition, please ensure you answer the following questions:
What did you learn about the data?
How did you answer the questions?
How can you justify your answers?