목록Data Science/Machine Learning (20)
Code&Data Insights
[Machine Learning] Regularization - Early Stopping | Weight Decay | Dropout
Regularization : prevent overfitting and improve the generalization of a model - It introduces additional constraints or penalties into the model training process to discourage the model from becoming too complex. - It aims to strike a balance between fitting the training data well and maintaining simplicity in the model Early Stopping : monitor the performance after each epoch on the validation..
[Machine Learning] Generalization, Capacity, Overfitting, Underfitting
Generalization : ability of a machine learning algorithm(model) to perform well on unseen data Training Loss : loss function, computed with training set Test Loss : loss function, computed with test set => More training set leads better generalization! Capacity : Underfittingand Overfittingare connected to the capacity of the model. capacity(= representational capacity) : attempts to quantify ho..
[Machine Learning] Hyper parameters | Feature Importance | Gini Impurity | Mean Decrease in Impurity
Hyper parameters Hyper parameters : configuration values used to control and tune the behavior of machine learning algorithms and models. These parameters have a significant impact on the model's training process and performance. => Properly setting hyperparameters can optimize the model's performance and prevent overfitting. => Model Performance Optimization, Preventing Overfitting, Saving Trai..
Evaluation | Classification Report - Accuracy, Precision, Recall, F1 Score
분류 예측시 confusion matrix로 TN, FN, TP, FP 를 표현 할 수 있음. 앞이 실제 값, 뒤가 예측값!!! ( FN - N 예측값 : Negative) TN - 실제 데이터셋 Negative(0), 예측 값 Negative(0) => True!!! TP - 실제 데이터셋 Positive(1), 예측 값 Positive(1) => True!!! FP - 실제 데이터셋 Negative(0), 예측 값 Positive(1) => False FN - 실제 데이터셋 Positive(1), 예측 값 Negative(0) => False 정확도 Accuracy rate - 예측 결과와 실제 값이 동일한 건수 계산? (TN + TP) / (TN + FN+TP+FP) 정밀도 Precision - 데..
[Machine Learning] Model Selection - K-Fold Cross Validation | Grid Search
[ K-Fold Cross Validation ] K-Fold Cross Validation : a validation technique used to evaluate the performance of a model. It involves dividing the given dataset into K different subsets (or folds), sequentially using each subset as the validation dataset, and using the remaining subsets as the training dataset. [ Grid Search ] Grid Search : Grid Search method involves training and evaluating the..
[Machine Learning] Dimensionality Reduction - Feature Extraction | PCA | LDA
[ Dimensionality Reduction ] Feature Selection - a process in machine learning and data analysis where a subset of relevant features (variables or attributes) is selected from a larger set of available features. - to improve the model's performance by reducing overfitting, improving interpretability, and enhancing computational efficiency. Feature Extraction - a process in machine learning and d..
[Machine Learning] Natural Language Processing(NLP) | Bag-Of-Words Model
[ Natural Language Processing(NLP) ] Natural Language Processing(NLP) : Natural Language Processing involves the development of algorithms and models that allow computers to comprehend, generate, and interpret sentences, as well as extract meaning from text. => NLP Process (1) text data is collected and preprocessed. During this stage, the text is segmented into sentences or words through tokeni..
[Machine Learning] Reinforcement Learning - Upper Confidence Bound(UCB) | Thompson Sampling
[ Reinforcement Learning ] Reinforcement Learning : reinforcement learning is a field of machine learning where a computer program, known as an agent, learns and improves gradually through experience while performing tasks in a specific environment. => Agent interacts with the environment, perceives its current state, selects and executes actions, and receives rewards. => During this learning pr..