jongwoncode / kaggle

participated in kaggle and dacon competition

💬 Kaggle and Dacon Competition Repository

💬 Repository 소개

• 주로 과거의 대회들을 중심으로 참여하여 공부한 기술들을 사용하는 것을 목적으로 하고 있습니다.

💬 Kaggle & Dacon & Programmers 대회 참여 list

👻[Kaggle] Quora Quetions Pairs

🥰사용 기술

• EDA : 통계량, Log-histogram, box-plot, wordcloud,barchart
• Preprocessing : 정규 표현식, Down Sampling, stopwords, deconstradictions
• Word Representation : TF-IDF, Word2Vec, TF-IDFweighted w2v model
• Dimension Reduction : SVM
• Scaler : Standard Scaler
• Parameter Search : Grid Search CV
• Model Normalizing Methd : Batch Normalization, Dropout
• Model : Logistic Regression, Random Forest,XGBoost, Decision Tree, Wide and Deep
• Stacking : stacking 3 model

👻[Programmers] Art Image Classifcation

🥰사용 기술

• Data Preprocessing : Resizing, Rescaling
• Data Augmentation : Horizontal Flip, Rotation, CutMix
• Train Validation Split : Stratified Classification
• Model : Transfer Learning, EfficientNet, Xception
• Stacking : Logistic Regression
• Optimizer : Adam, Cosine annealing learning rate scheduler
• Metric : Accuracy

💬🥰😎🤑👻🥳😄👩‍🏫🤡👨‍🏫🤑🤬😬🙂