Projects on Yixian Wang

Android App: Business Management & Accountant

Wed, 25 Oct 2023 00:15:21 -0400

Designed the app for my parents’ real daily business
Features: cloud-hosted database, interactive charts, automatic accountant, and PDF generator
Backend jobs: firebase for authentication, firestore for data and SQL query
Architecture & Language: MVVM, Kotlin
Third-party libraries: MPAndroid, Itextpdf, Android-pdf-viewer

Fri, 24 Dec 2021 23:24:06 -0400

Implemented astrophysical simulation solved N-body problem using Barnes-Hut algorithm with MPICH by C++
Programmed OpenGL to visualize the movement of the bodies in the domain by C++
Analyzed performance with the number of bodies, processors, timesteps, iterations

Sun, 24 Oct 2021 17:17:01 -0400

Implemented concurrency programming model to compute BST(binary search tree) equivalence with Go
Implemented channels, go-routines, and signaling with Go
Programmed threads to parallelize hash operations with Go
Assembled a concurrent buffer to secure communication among threads
Analyzed performance among all the implementations

Sun, 24 Oct 2021 15:21:05 -0400

Implemented base sequential version of K-Means algorithm on CPU by C++
Implemented first parallel version of K-Means with Thrust primitives and a GroupBy-Aggregate by C++
Implemented second parallel version of K-Means with CUDA by C++
Implemented third parallel version of K-Means with CUDA on Shared Memory by C++
Analyzed speedup among all the implementations

Fri, 24 Sep 2021 18:44:08 -0400

Implemented base sequential version of work-efficient parallel prefix sum algorithm by C++
Implemented parallel and barrier versions of work-efficient parallel prefix sum with POSIX thread (pthread) by C++
Analyzed speedup among the all implantations with respect to the number of threads and data size

Fri, 25 Jun 2021 02:08:49 -0400

Designed deep networks for a racing simulator, SuperTuxKart with Pytorch
Trained linear model and multi-layer perceptron model to classify images from SuperTuxKart
Trained a convolutional network to classify images from SuperTuxKart
Built classification network fully convolutional and solved a semantic labeling task (labeling every pixel in the image)
Implemented an object detector
Trained a CNN to do vision-based self-driving in SuperTuxKart
Programmed a SuperTuxKart ice-hockey player (AI that plays ice-hockey)

Sun, 25 Apr 2021 02:18:07 -0400

Implemented an Encoder-Decoder model for semantic parsing with Pytorch
Implemented a decoder by using LTSM whose output is passed to a feedforward layer and a softmax over the vocabulary
Added attention mechanisms to the model to make it more powerful and faster to train