代写 CE 451/551、代做 Python编程语言

日期：2025-04-19 10:42

CE 451/551 – Computer-Aided Research in the Chemical and Materials Sciences: Homework #18 (Bonus #6)

(Due: Tuesday, 6 May 2023, 5:00pm)

The goal of this assignment is to improve your understanding of artificial neural networks (ANNs) and their implementation by manipulating a given ANN Python code. Your task is to take the final ANN code from Stephen Welch’s “Neural Networks Demystified” presentation (i.e., Jupyter notebook Part 7, which you can download here: , and rewrite it in the following sense:

1) Replace the original data for “X” and “y” with data from our sklearn tutorial, i.e., “X” should be the data given in the file molorg_features.csv (specifically, the columns 'SpPos_B(p)' and 'VE2sign_L'), and “y” should be the data given in the file molorg_pol.csv. Note that you will have to write a little code to read in this data and bring it into the numpy array format required by Welch’s Jupyter Notebook. You will also have to adjust the labeling and make any other changes required for this notebook to run. (Bonus: 5 points.)

2) After completing (1), increase the number of neurons from 3 to 10 and compare your results with those of (1). (Bonus: 5 points.)

3) After completing (2), increase the number of hidden layers from 1 to 2 and thus make this a deep neural network. Compare your results with those of (1) and (2). (Bonus: 10 points.)

Please write up a short summary (1 paragraph for each part of the assignment) in which you describe the changes you made and discuss the results you obtain. Your discussion should rationalize your findings.

Please submit your zipped Jupyter notebook file to me via email – do not just copy your source into the email. Your submission will be graded based on the quality of your implementation (including efficiency, documentation, readability, clarity), and whether your program actually works or not. As part of the evaluation, I have to be able to execute the file you send me using the current Anaconda version. Please provide your writeup in a doc or pdf file. Use Redmine to plan and document your development activities. Please submit your zipped notebook back to me via email. Please use the following email subject line “CE 451/551 HW18 submission by <your name>” .

Tip 1: Stephen Welch’s “Neural Networks Demystified” presentation/youtube video will show you where your code will need to be rewritten.

Tip 2: If you have trouble getting your code to run, try to google the error message you get. They will likely lead you to a useful stackoverflow page that will help you solve the problem.

Note: Your code should be adequately commented, so that I know what you are doing.

Note: You may interact on technical questions with your classmates, but every student has to submit an individual solution. No two scripts/programs can be alike.

Reminder of the rules:

CE 451/551 has no exams. Grades are primarily determined on the basis of ten graded assignments, each contributing 10% to the final grade. Each graded assignment (and thus the final grade) has a 0-100 points scale. In addition to the graded assignments, there are five pass/fail assignments covering essential content every student has to master to succeed in this class. Failure to complete the latter results in a predetermined number of penalty points that are deducted from the final grade. (Note that you do not literally fail the class if you do not complete a pass/fail assignment, however, the penalty points will have a severely negative impact on your final grade.) In

 1

addition to the mandatory assignments, there are seven voluntary bonus assignments. Completion of the latter is rewarded with a predetermined number of bonus points that are added to the final grade. Active in-class participation throughout the semester is rewarded. It is recorded after each class and tallied for each student at the end of the semester. Up to 5 bonus points may be added to the final grade for outstanding and exceptional contributions throughout the course. Note that these bonus points are not required to receive a perfect grade, but that they can compensate for points lost in other places.

The overall letter grades are based on 5-points brackets:

Points 100-96 95-91 90-86 85-81 80-76 75-71 70-66 65-61 60-56 55-51 50-0 Letter A A− B+ B B− C+ C C− D+ D F

In past years, the course average has been around an A−/B+. I reserve the right to curve the grades, should the need arise (however, this has never been necessary so far). I will provide intermediate grades and/or grade projections at regular intervals throughout the semester and you are encouraged to proactively request updates as well. In addition, I am providing the grade_tinker.xlsx spreadsheet on UB Learns with which you can keep track of your standing and assess different grade scenarios.

Good grades are very achievable, and they typically strongly correlate with individual students’ engagement.

The performance expectations that form the basis for the grading of the assignments differ for the graduate vs undergraduate section of the course. In particular for the coding projects, the graduate students are expected to deliver more extensive and technically advanced products that tie into their research work. The grading scale is adjusted correspondingly.

Assignments come in different shapes and forms, including traditional homework, coding tasks, projects, and reports – either for groups or on an individual basis. Assignments and due dates are posted on UB Learns (or given individually), and instructions for electronic submissions via email or UB Learns are provided. Please follow these instructions and the naming conventions for the submissions exactly and submit each assignment individually (rather than in a bundle). Unless stated otherwise, there is a 20-point penalty for late assignments, and additional 20 points are deducted after each additional 24h (given a good reason, I may reduce or wave a late penalty). Late penalties start to apply once the assignment deadline passes. I am open to discussing no-penalty extensions of deadlines, hardship exceptions, and accommodations if warranted.

Late >0h, <24h >24h, <48h >48h, <72h >72h, <96h >96h

Penalty −20 pts −40 pts −60 pts −80 pts −100 pts

Failure to follow all assignment instructions will result in penalty points. Work that is disorganized, unclear, illegible, or otherwise unprofessionally presented may have points deducted or be returned without a grade at my discretion. Subsequent resubmission may be considered late. To receive full credit, you must show all the logical steps of your work. Extra points may be awarded for particularly original solutions, and penalty points may be deducted for outrageously wrong or obviously nonsensical answers (i.e., you may want to leave blanks rather than submit wild guesses). Precision is a virtue, so please avoid fluffed out and waffling answers. Make-up or do-over assignments are generally not offered. These rules may seem strict, but there are many assignments to grade and failure to comply with the rules makes life unnecessarily hard for me and the graders.

Mistakes happen and if you think that you unjustly lost points on an assignment, please write a few sentences explaining your position and making a substantive case for a revision. Send this response to me for review. I will evaluate if you make a good argument, and your claim has merit and correct the corresponding grade if warranted. You can submit claims via email, in class, or leave them in my mailbox in 308 Furnas Hall. I will only accept regrade requests within a week after an assignment is returned, so that issues can be resolved in a timely fashion. Note that we have procedures in place to identify attempts at tempering.

While you are encouraged to collaborate and exchange ideas with your classmates, your work has to be individually and independently written up (unless stated otherwise). No two students' solutions can be identical, nor can they be direct or disguised copies from solution manuals, the internet, or similar sources. Failure to comply constitutes academic dishonesty and carries penalties as discussed below.

                                                            2

Academic integrity

Breaches of academic integrity (e.g., plagiarism, cheating, purchasing or selling of class notes or assignment solutions) are unacceptable and will result in a failing grade for the particular assignment and/or for the entire course. It’s not right and it’s not worth it! Academic integrity is a fundamental university value. Through the honest completion of academic work, students sustain the integrity of the university and of themselves while facilitating the university's imperative for the transmission of knowledge and culture based upon the generation of new and innovative ideas. It is expected that you behave in an honorable and respectful way as you learn and share ideas. To summarize UB’s policy on dishonesty: A student will not present, as his or her own, the work of another, or any work that has not been honestly performed; will not take any examination by improper means and will not aid and abet another in any dishonesty. Please consult UB’s Student Code of Conduct and UB’s Undergraduate Academic Integrity Policy at:

Note: Cheating on an assignment or exam does you no good, and dishonesty reflects poorly on your character. Any incident of academic misconduct, regardless of severity, will be brought to UB’s Office of Academic Integrity. At a very practical level, you should ask yourself how you might expect a reference from faculty members if you have been cheating in their classes. I had to punish offenders before, and it is tough on everyone involved. PLEASE, spare yourself and me this painful situation! You have been warned!

Copyrighted course materials

Course materials (lecture/recitation slides, recordings, files, assignments, master solutions, examples, etc) are for course purposes only, they are copyrighted, and they may not be shared outside this class without my prior written permission. In particular, they may not be shared on Course Hero, Chegg, or similar platforms, or with anyone outside the class. Please note that my course material may contain individualized hidden watermarks and that UB can monitor network access to cheating platforms. Failure to comply constitutes academic dishonesty and carries penalties as discussed above.

Please use your @buffalo.edu account when you communicate via email.

      3