Only the expert who completed my last assignment will be able to complete this as it is a continuation of the last assignment, I believe the name was Kunal. I will attach the previous assignment(Prev.pdf), the current assignment (Current.pdf), and the format the paper must be written in(How to write P1.pdf) to this.
.
Mini Proj 4 Mini Proj 4 Noah Hardin Data Preparation Display some sample images Data Preparation Continued Import pretrained network and tune dataset Replace Final Layers Train Model Augmentations Setting training parameters Cross Validation Network current You will be writing a short, academic technical paper formally explaining your face recognition assignment. The intent is to find out how well you can understand, analyze, and explain your project results and observations using the theories taught during this course. In a nutshell, this should read like a mini IEEE journal paper. Length: 1-2 single spaced pages of dense text for the main body (12 point, single space, any san serif font) not counting extensive white space, references, figures, tables, formulas (they will go into the appendix). Provide numbered captions for all your figures (on bottom), tables (centered on top), and formulas (to their right) and reference them accordingly in the text. Paper Structure: You need to expand upon four simple but critical points (underlined): 1-What (precise problem statement) 2-Why (importance of the subject matter, AND importance of your approach) 3-How (your data and methods, described to the level that your experiment is reproducible by your audience) 4-Discuss and justify your results--THIS IS THE MOST IMPORTANT PART P1-Title: Use a relevant title such as “Using Transfer Learning and Deep Neural Networks for Face Recognition” (you may change this suggested titles if you wish) ECE416/5316//CS5590NN FS 2021 Final Paper Guide (Take Home Final) P2-Your name and declaration of sole authorship (DON’T FORGET THIS!): Assert in writing that you wrote the whole paper all by yourself, in your own words, and that all external knowledge and literature has been referenced properly in the text. P3-Abstract: This should be a snap-shot of the whole paper, in 200 words or less. Start with a couple of sentences summarizing S1 (please see below), a short but precise summery of data and the utilized computational methods (e.g. different models used), and then finish by the most representative results (including accuracy numbers) and one sentence on your main conclusion. S1-Introduction (What and Why) A short description of the dataset, followed by a couple of paragraphs about importance of deep learning and its applications in face recognition. S2-Methods (How) Precisely and concisely describe data preparation and classification methods you used. Write at least one paragraph for each method, accompanied by corresponding reference. Put formulas in the appendix. S3-Results and Discussions (Analysis and Justification of Results) This is the main part of your paper and should constitute about 40%-50% of its length. Cogently write down, discuss, and justify the results of your experiments and observations from all the cases you tried for your project: what types and ranges of classification parameters did you try in each case, and why? What trends did you see? You need to justify and explain everything referring to the basic concepts and theory taught in class, citing relevant references from your textbook(s) (again, use page numbers in addition to the standard reference). Elucidate your point by referring to corresponding enumerated tables, figures, and formulas in the appendix. S4-Conclusion Start by a summary of your work, and finish by your major conclusions and take home messages (e.g. on the capabilities and limitations of the used methods given your project experience). S5-References Examples: [1] A. Subasi, "EEG signal classification using wavelet feature extraction and a mixture of expert model," Expert Systems with Applications, vol. 32, pp. 1084-1093, 2007. [2] J. C. Príncipe, et al., Neural and adaptive systems: fundamentals through simulations. New York: Wiley, 1999. [3] I. Goodfellow, et al. Deep Learning. Cambridge: MIT Press, 2016. ECE416/5316//CS5590NN FS 2021 Final Paper Guide (Take Home Final) S6-Appendices: All the tables and figures (along with their captions), as well as the formulas you referenced in the body of your paper go here. You can reuse material from your submitted slide deck for the project. http://mail.philadelphia.edu.jo.PDF Fourteen Steps to a Clearly Written Technical Paper by R. T. Compton, Jr. A technical paper will usually have four sections. The purpose of each of these sections is as follows: Section I: Introduction The introduction should do the following: 1. Open up the subject. (The subject will be electromagnetic fields in cylindrical dielectric geometrics, adaptive arrays in packet radio, or whatever.) 2. Survey past work relevant to this paper. 3. Describe the problem addressed in this paper, and show how this work relates to, or augments, previous work. 4. Describe the assumptions made in general terms, and state what results have been obtained. (This gives the reader an initial overview of what problem is addressed in the paper and what has been achieved.) 5. Overview the contents of the paper. (LSection II contains our formulation of the problem. Section III contains the experimental data...N) Section II: Formulation of the Problem This section should do three things: 1. Define the problem to be considered in detail. Typically this section might begin with something like: LConsider a packet radio system consisting of a single central repeater surrounded by user terminals. Each user transmits packets to the central repeater using a slotted ALOHA protocol [1]. The transmissions from all users are assumed to be on the same frequency...N The discussion should proceed in this way until the problem is completely defined. 2. Define all terminology and notation used. Usually the terminology and notation are defined along with the problem itself. 3. Develop the equations on which your results will be based and/or describe any experimental systems. Section III: Results This section presents the detailed results you have obtained. If the paper is theoretical, you will probably show curves obtained from your equations. If the paper is experimental, you will be presenting curves showing the measurement results. In order to choose the proper curves to present, you must first be clear what point you are trying to convey to the reader. The curves can then be chosen to illustrate this point. Whether your paper is theoretical or experimental, you must provide a careful interpretation of what your results mean and why they behave as they do. Section IV: Conclusion This section should summarize what has been accomplished in the paper. Many readers will read only the Introduction and Conclusion of your paper. The Conclusion should be written so they can be understood by someone who has not read the main work of the paper. This is the common format for an engineering paper. Of course, the names of the sections may differ slightly from those above, but the purpose of each section will usually be as described. Some papers include additional sections or differ from the above outline in one way or another. However, the outline just presented is a good starting point for writing a technical paper. To write your paper, you should proceed as follows: Step 1: Start by writing a complete first draft of your paper, except for the Introduction and Conclusion. (It is easiest to leave the Introduction and Conclusion until after the main body of the paper is written.) In writing your paper, keep the following in mind: 1. You must always present the big picture first and then work towards the details. The other way around will not work. This is especially true in the beginning of Section II, where you are explaining the problem you are studying. 2. If you get stuck and cannot figure out how to explain something, a useful trick is to imagine that you are telling a very good friend what you are working on: just put down the words as you would say them to your friend. In writing your first draft, do not worry if the wording is not perfect. Polishing the document comes later. When you are finished with your first draft, put it away for a couple of days before you begin Step 2. Step 2: Make sure the ideas in the paper are in the right order. If not, move blocks of the paper around with your text editor until they are. Ask yourself: LCan the reader understand every passage strictly from the material up to that point?N If not, add material or move ideas around. Make sure there are not gaps in your logical arguments, and make sure you are not implicitly assuming that the reader understands something needed to follow your arguments, even though you have not stated it. The reader probably understands less than you think. Step 3: Work on the transitions between ideas. Make sure that at each stage the reader has a roadmap of where he or she is going. The reader must be able to see the big picture. At the beginning of each section, make clear to the reader in advance what the purpose of that section will be and how that section relates to the preceding material. At the end of each section, you may also want to remind the reader that you have now completed what you set out to do in that section. Then point out what the purpose of the next section will be, and so forth. These connecting statements are called transitions. The reader must always be able to see where you are going and why and how far you have progressed. Step 4: Check each paragraph for unity. Each paragraph should have one main point. Usually the central point of each paragraph is stated in a topical sentence at the beginning of the paragraph, but not always. You should not mix different ideas together in the same paragraph. If you are having trouble getting a certain section of your paper to sound right, go through that section one paragraph at a time and ask yourself what the main point of each paragraph is. Foggy writing is often due to mixed-up paragraphs. Step 5: Work on the sentences to reduce the fog index. The Fog Index F is defined as F = 0.4(L + P), where L is the average number of words per sentence and P is the average number of polysyllables per 100 words of text [1]. (A polysyllable is a word with three or more syllables.) To evaluate the Fog Index for your paper, count the number of words per sentence and the number of polysyllables per 100 words for a representative portion of your paper five or six hundred words long. Ideally, you should strive for a fog index less than 10. In technical writing, it is sometimes difficult to get the Fog Index below 10, but a Fog Index above 15 is a warning that your material will be very hard for a reader to follow. Consider the following examples taken from typical office memos. Note that both memos say the same thing. Fog Index = 35: LIn order to eliminate the possibility of errors occurring in the time charges relating