{"id":19515682,"url":"https://github.com/presidentsam100/sam-mit-future-makers","last_synced_at":"2025-06-24T06:33:10.122Z","repository":{"id":162779622,"uuid":"383841241","full_name":"PresidentSam100/Sam-MIT-Future-Makers","owner":"PresidentSam100","description":"MIT FutureMakers curriculum projects.","archived":false,"fork":false,"pushed_at":"2021-08-13T15:42:43.000Z","size":2021,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-01-08T12:35:52.585Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/PresidentSam100.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2021-07-07T15:16:49.000Z","updated_at":"2021-08-13T15:42:46.000Z","dependencies_parsed_at":"2023-05-27T21:15:33.968Z","dependency_job_id":null,"html_url":"https://github.com/PresidentSam100/Sam-MIT-Future-Makers","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/PresidentSam100%2FSam-MIT-Future-Makers","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/PresidentSam100%2FSam-MIT-Future-Makers/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/PresidentSam100%2FSam-MIT-Future-Makers/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/PresidentSam100%2FSam-MIT-Future-Makers/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/PresidentSam100","download_url":"https://codeload.github.com/PresidentSam100/Sam-MIT-Future-Makers/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":240766674,"owners_count":19854114,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-11-10T23:40:43.890Z","updated_at":"2025-02-25T23:43:31.194Z","avatar_url":"https://github.com/PresidentSam100.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Sam-MIT-Future-Makers\n\n- [Deep Learning Curriculum Responses](#deep-learning-curriculum-responses)\n  - [Week 1](#week-1)\n  - [Week 2](#week-2)\n  - [Week 3](#week-3)\n  - [Week 4](#week-4)\n- [Create-a-thon Progress](#create-a-thon-progress)\n  - [Week 5](#week-5)\n  - [Week 6](#week-6)\n\n## Deep Learning Curriculum Responses\n\n### Week 1\n\n#### Day 0 (July 5th, 2021)\n\nReflect:\nI was introduced to the program and had to setup a Github account and downloaded other files that included the curriculum and the schedule for the program. I also took part in a get to know each other day where all the students of the program would introduce themselves and get to know everyone before it officially starts.\n\n#### Day 1 (July 6th, 2021)\nReflect:\nI'm excited to expand my skills in Python and use some AI knowledge to work on a project with my group. My group I included mentor Kevin Walsh and teammates Diya Jim, Alyssa Tang, Sahana Sreeram, Jack Tan, and myself. I hope to make use of these skills when developing projects for the Create-A-Thon during the summer.\n\nLessons:\nI don't have any specific skills that I want to learn, but I want to get an idea of the AI field and getting to know about this during the summer.\n\n#### Day 2 (July 7th, 2021)\nReflect:\nI learned so much from the Leadership Seminar from David Kong and learned how to formulate and share a story by describing the experience I had as well as formulating it in a way that allows others to feel the same way I did during that experience. \n\nLesson:\nMore specifically, I recognized that when talking about oneself, I should talk about a challenge that I experienced and how I solved it, and something that I took away from the experience. By doing so, it allows me to recognize that any contribution, even the smallest, can have a significant impact, and help engage others in getting involved too.\n\n#### Day 3 (July 8th, 2021)\nReflect:\nI was able to hear some challenging experiences that some might face during their lives and learned how to come up with a plan or solution to help accommodate for those challenges and to discuss those solutions with other peers.\n\nLesson:\nSupervised learning and unsupervised learning are two specific types of machine learning, but the key difference between them lies in whether you have to \"supervise\" them. In particular, supervised learning takes in data and learns from that feedback such as classification models, but unsupervised learning only learns from input data and uses patterns such as a cluster model. Even though Scikit-learn can't produce visualizations on its own, other libraries such as matplotlib can help with that.\n\n#### Day 4 (July 9th, 2021)\nReflect:\nI learned about the differences between Deep Learning (DL) and Machine Learning (ML), and I also learned about CPUs, and other processing units I had never even heard of before such as TPUs and GPUs.\n\nLesson:\nOne real-world problem that is common especially with technology is the hacking of passwords especially if similar passwords are used between accounts associated with the same user. This is a problem because by doing so, certain accounts are more vulnerable and significant amounts of crucial data and information can be leaked.\n\nDataset:\nOne dataset I found from Kaggle demonstrates how poor password safety can be a problem (https://www.kaggle.com/wjburns/common-password-list-rockyoutxt)\n\n### Week 2\n\n#### Day 7 (July 12th, 2021)\nReflect:\nI learned about Neural Networks and its building blocks and how it relates to the human brain and how it can be used to predict how most humans would try and behave. I also learned more about the Tensorflow, Keras, and Pandas libraries.\n\nLesson:\nTensors store data in a way similar to matrices but can be any dimension, unlike matrices which are 2 dimensional. Tensors can be scalar (0 dimensions), vector (1 dimension), matrix (2 dimensions), or anything greater. However, most tensors are generally used with more dimensions.\n\nObservation:\nSomething interesting I discovered is that TensorFlow is constantly updating and new features can make older ones obsolete. In addition, there were many different functions that each library can do such as creating a graph of data or creating a simple neural network.\n\n#### Day 8 (July 13th, 2021)\nReflect:\nI listened to a talk given by Eric Weber discussing the importance of data and understood the connection between data science and machine learning where one possilbly work well on its own but could be enhanced by the other.\n\nLesson: I learned about Neural Networks and its building blocks and how it relates to the human brain and how it can be used to predict how most humans would try and behave. Neural networks\n\n#### Day 9 (July 14th, 2021)\nReflect:\nActivation functions can be used to find features that are not necessarily given within data. Rectified Linear Unit Layers are nonlinear functions with no parameters given which results in an output that resembles a retified feature map.\n\nLesson:\nI learned about convolutional neural networks (CNN) and how it's used for images since each image is a matrix of pixel values and was able to dive into a topic that I had been unfamiliar with at first.\n\n#### Day 10 (July 15th, 2021)\nReflection: \nHow do you think Machine Learning or AI concepts were utilized in the design of this game?\nI feel that Machine Learning and AI concepts can be used in any game such as Pong (https://www.ponggame.org) because the computer controlled paddle will move without any human commands and will move to try and get the ping pong ball and can be set to different levels of difficulty which can make the AI more accurate or play in a way that tries to win rather than not lose.\n\nCan you give a real-world example of a biased machine learning model, and share your ideas on how you make this model more fair, inclusive, and equitable? Please reflect on why you selected this specific biased model.\nOne real-world machine learning model that I believe has a bit of bias is Youtube or at least its algorithm in recommending certain videos to specific users. Part of the reason why I feel that there is some bias in this model is because it tends to recommend certain videos that it wants its users to watch and would stay on the website for much longer so it tends to recommend more of those entertaining but somewhat addicting videos rather than those that come out of our interest and past watches. One way this model could be made more fair is to allow for certain options of pausing search history or only allowing specific types of videos to be recommended to be made more clear because while such features do exist, they are hardly put to use by a large amount of users due to a lack of awareness so making those features more obvious would help greatly.\n\nLesson:\nI learned about the types of biases and considerations for even the most non-biased of AI algorithms because they might favor certain results over another and how we can spot and respond to such biases.\n\n#### Day 11 (July 16th, 2021)\nReflection:\nI learned about the types of biases and considerations for even the most non-biased of AI algorithms because they might favor certain results over another and how we can spot and respond to such biases.\n\nLesson:\nI learned about the differences between a convolutional neural network (CNN) and a fully connected neural network (FCNN). A CNN can consist of a pooling layer or a fully connected layer while a FCNN consists of a group of such fully connected layers. CNNs can be used for computer vision in image searching and facial recignition while FCNNs have a more broad applicability.\n\n### Week 3\n\n#### Day 14 (July 19th, 2021)\nReflection: I found that several math topics that were learned in high school or even college could be applied to machine learning in some way especially those used in statistics with the loss functions.\n\nLesson: I learned about loss functions which primarily aid in portraying the error of the model between the predicted value and actual value. I also learned that the mean squared error loss is similar to a least-squares regression line in statistics as they both use some methods in trying to minimize the effect of incorrect or otherwise less accurate predictions for data.\n\n#### Day 15 (July 20th, 2021)\nReflection: I listened to vice president of engineering at Principles, Daniel Ranallo, discuss his view on the importance of Artificial Intelligence and Machine Learning and realized the possibilities and growth of such a field in the future where technology is becoming increasingly important and will guide much of our lives.\n\nLesson: I learned about ReLU functions which are mainly used for speech recognition and natural language processing. ReLU functions are more efficient and faster than other activiation functions and are similar to the way neurons work in the brain.\n\n#### Day 16 (July 21th, 2021)\nReflection: Factors such as exclusion, over-representation, and biased labeling all can influence the predictions that a machine learning model makes. Therefore, it is important to consider the ethics when designing a model since no such model will be free of any bias or have any objective truth.\n\nLesson: I learned about the ethics of AI and why certain companies designed their AI models in their particular manner. I also looked back on some of the previous curriculum topics to understand how they relate to AI ethics.\n\n#### Day 17 (July 22nd, 2021)\nReflection: Senior research scientist, Jennifer Chu-Carrol, gave a talk on interactive and collaborative problem solving. The main takeaways were that communication with the team was crucial in having the final project to work and that an effective plan before designing the actual product is essentially mandatory.\n\nLesson: I learned about image classification and some of the tools that could be used but also encountered and learned some significant challenges that were possible with image classification such as the illumination or viewpoint variation that would make it harder.\n\n#### Day 18 (July 23rd, 2021)\nReflection: I learned some ways that overfitting could be harmful such as when the data is merely memorized and not effectively used to make accurate predictions.\n\nLesson: I learned about overfitting which is when the model is trained too much and the error cannot be narrowed down or analyzed with high accuracy. \n\n### Week 4\n\n#### Day 21 (July 26th, 2021)\nReflection: I modified and improved on some of my designs of previous models to increase their accuracy or make their learning techniques more efficient. I also began experimenting on some of the curriculum topics to have an idea on which topics would be best for the create-a-thon.\n\nLesson: I learned about autoencoders which are neural networks that learn data unsupervised. They can also be used to correct certain data that has been corrupted or even denoise images. For example, it could show a very blurry letter or number and then determine the character that was blurred with high accuracy.\n\n#### Day 22 (July 27th, 2021)\nReflection: I was able to get a grasp on building my profile and resume and I also was able to learn networking skills to help let others know about my skills and to get more involved to apply some of my skills in a real-world environment where I would be working on a project with a team.\n\nLesson: I learned about affective computing which attempts to use machine learning algorithms to identify and analyze human emotions. I feel that this would be very important in my future projects because emotions are a key feature in recognizing the user and would feel that it could help others who are not as good in processing or understanding certain emotions.\n\n#### Day 23 (July 28th, 2021)\nReflection: I learned about Natural Language Processing (NLP) which is the study of comptational treatment of the human language. I was also able to improve upon some of the code that I had written in the past few days in preparation for the upcoming Create-a-thon.\n\nLesson: NLP has many practical applications including speech recognition, translation between languages, and prediction of future words. In order to preprocess data, NPL can convert text into an encoded vetor or can geneerate clean ASCII text or can use lemmataization.\n\n#### Day 24 (July 29th, 2021)\nReflection: I listened to Erica Greene and her webinar on Machine Learning and its increasing use in tech companies. I also explored the topic of Computer Vision and used that topic on that days's project where I analyzed images and determined its meaning and application in the field of AI.\n\nLesson: I learned how machines can read one's emotions by simply looking at a person's face and listened to a TED Talk given by Rana El Kaliouby on implementing emotion detection in my project.\n\n#### Day 25 (July 30th, 2021)\nReflection: I gained an understanding of the entrepreneurial mindset for the Create-a-thon that will take place within the last two weeks of this program. I listened to speaker Carly Chase and her advice on taking such an active role by coming up with a plan, and determining many steps and goals before going on to design the actual product.\n\nLesson: Even though I didn't learn any new deep learning concepts that weren't covered in the previous days, I had the time to use this day to catch up on any past assignments and improve my Github and make any necessary updates to my past code right before the Create-a-thon.\n\n## Create-a-thon Progress\n\n### Week 5\n\n#### Day 28 (August 2nd, 2021)\nProgress: My team brainstormed ideas for the topic of our deep learning model.\n\nComputational Talk: I listened to MIT student Nicole Pang give a talk regarding implementing what the team has learned during your Create-a-thon project ideation, planning and execution.\n\n#### Day 29 (August 3rd, 2021)\n\n#### Day 30 (August 4th, 2021)\n\n#### Day 31 (August 5th, 2021)\n\n#### Day 32 (August 6th, 2021)\n\n### Week 6\n\n#### Day 35 (August 9th, 2021)\n\n#### Day 36 (August 10th, 2021)\n\n#### Day 37 (August 11th, 2021)\n\n#### Day 38 (August 12th, 2021)\n\n#### Day 39 (August 13th, 2021)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpresidentsam100%2Fsam-mit-future-makers","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpresidentsam100%2Fsam-mit-future-makers","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpresidentsam100%2Fsam-mit-future-makers/lists"}