In this post I’ll be discussing my final project for ECE 685: Deep Learning, a course I took at Duke University. For this project, Jiawei Chen and I proposed a two-stage algorithm to improve classification accuracy. You can find my Github repository for this blog here. Convolutional Neural Networks (CNNs) have been used on 3D point clouds for object classification. However, due to the nature of the CNNs, classifiers, especially those CNN-based classifiers, are usually confused about objects that look alike.

Continue reading

1. Background COVID-19, a disease caused by a new type of coronavirus, has become a major global human threat that has turned into a pandemic. During this period, the role of the immune system has attracted people's attention. Some researchers identified some important nutritional considerations for the prevention and management of COVID-19 diseases (Yasemin Ipek Ayseli et al., 2020). Based on these studies, some "experts" and articles have urged people to buy supplements or eat particular foods to enhance their immune system.

Continue reading

In this post, I would like to share my implementation of several famous first-order optimization methods. I know that these methods have been implemented very well in many packages, but I hope my implementation can help you understand the ideas behind it. Suppose we have \(N\) data examples and the parameters \(\mathbf{w} \in \mathcal{R}^D\). For convenience, I first write a class named optimizer. class optimizer: def __init__(self): pass def set_param(self, parameters): self.

Continue reading

As we all know, insects are a major factor in the world's agricultural economy. Therefore, it is particularly important to prevent and control agricultural insects by using procedures such as dynamic surveys and real-time monitoring systems for insect population management. However, there are many insects in the farmland, and it takes a lot of time to be manually classified by insect experts. Since people without the knowledge of entomology cannot distinguish the types of insects, it is necessary to develop more rapid and effective methods to solve this problem.

Continue reading

I believe that for graduate students, what to do after graduation may not have been decided yet. Is it a good choice to continue to study for a PhD, or is it a wise choice to find a job in the industry? What will doctoral degrees bring to students? Which field of doctoral degrees are more popular? To answer these questions, I found some historical data about doctorates awarded in the United States to display some information behind the data.

Continue reading

Star Wars API (SWAPI) is the world's first quantitative Star Wars data set that can be used for programming. The developers have aggregated multiple types of entity data involved in the Star Wars series of movies. It also provides a Python programming language package, swapi-python, which is built by the author of swapi, Paul Hallett. The six APIs correspond to six types of entities: Films：http://swapi.co/api/films/1 People：http://swapi.co/api/people/1 Starships：http://swapi.co/api/starships/1 Vehicles：http://swapi.co/api/vehicles/1 Species：http://swapi.

Continue reading

During my internship at Birch.ai, we want to test and improve the performance of our machine translation model. To achieve this, we first need to obtain a foreign language data set and its corresponding English data set, and hope that these data sets are translated by humans rather than machine translations. In this article, I would share my experience in scraping The New England Journal of Medicine from scratch using Python.

Continue reading