My code, data science, and machine learning practices.
In the car_rent_platform, it was my first time using Class in Python. The relevant code and material:
In the Cohorts of Songs project, I used Numpy, Pandas, Matplotlib, and Seaborn for EDA and Cluster Analysis. The relevant code, data, and material:
In the ScienceQtech project, I used MySQL to analyze data, such as creating an ER diagram, creating a view, querying using functions, and creating an index. The relevant code and material:
In the United States Airlines Analysis Capstone 2 project, the objective of this project is to identify the factors that contribute to avoidable flight delays and to build a model to predict if the flight will be delayed. I used Pandas, Numpy, BeautifulSoup, Matplotlib, and Seaborn for EDA and Analysis, Scikit-learn for DecisionTreeClassifier, and Xgboost in the project.
I studied and got a certificate for Machine Learning Specialization in Coursera, and the instructor is Andrew Ng.
The Fake News Classification aims to find out the fake news, it uses accumulative data, which could be used in news-based stock investment analysis. In the project, I compared the Sequential Model with Bi-Directional RNN and LSTM, the Decision Tree Model, and the Logistic Regression Model. I used Nltk, Tensorflow, Pandas, Matplotlib, Worldclud, Gensim, Keras, and Sklearn.
The relevant code:
Titanic Survivor Prediction with Logistic Regression and Naive Bayes.ipynb is used for Logistic Regression and Naive Bayes practices.
In the Deep Learning for Real Estate Price Prediction, I compared two different deep learning models: one used a limited number of features, while the other utilized an increased number of features. I used Pandas, Numpy, Matplotlib, Seaborn, Axes3D, Folium, and HeatMap for EDA and Analysis, TensorFlow and Keras for Deep Learning, and Adam in the project.
In the Build a Generative Adversarial Network, I used PyTorch to create the Generator Network and the Discriminator Network and to create a training loop to train the GAN model.
In the Building_a_Regression_Model_in_Keras, I used Keras to build a regression model to predict the compressive strength of different samples of concrete.
In the Final Proj-Classify Waste Products Using TL FT, I used TensorFlow, Keras, callbacks (for early stopping), and transfer learning of the pre-trained VGG-16 model.