동선생

Twelveth competition following Youhan Lee's curriculum. Natural Language Processing competition.First Kernel: Mercari Interactive EDA + Topic ModellingEDA kernel with matplotlib.Insight / Summary:1. Log transformation on target varOur response or target variables is the price we are suggesting to the Mercari's marketplace sellers. The median price of all the items in the training is about $267 b..

Fourteenth(Last) course following Youhan Lee's curriculum. Not competition.First Kernel: Novice to GrandmasterThe biggest problem that we might face is fake and bogus responses. As it is a survey, not everyone will answer with proper credentials, and thus I assume that there will be a lot many outlier. Second Kernel: What do Kagglers say about Data Science ?EDA Kernel with trying some prediction..

Thirteenth competition following Youhan Lee's curriculum. Natural Language Processing competition.First Kernel: [For Beginners] Tackling Toxic Using KerasKernel using keras LSTM.Insight / Summary:1. Checking null valuestrain.isnull().any(),test.isnull().any() 2. Tokenizationlist_tokenized_train = tokenizer.texts_to_sequences(list_sentences_train)list_tokenized_test = tokenizer.texts_to_sequences..

Eleventh competition following Youhan Lee's curriculum. Natural language processing competition. Spooky Author IdentificationShare code and discuss insights to identify horror authors from their writingswww.kaggle.comFirst Kernel: Spooky NLP and Topic Modelling tutorialTopic modeling: the process in which we try uncover abstract themes or "topics" based on the underlying documents and words in a..

Tenth competition following Youhan Lee's curriculum. Anomaly detection competition using tabular data. Credit Card Fraud DetectionAnonymized credit card transactions labeled as fraudulent or genuinewww.kaggle.comFirst Kernel: In depth skewed data classif. (93% recall acc now)Testing different methods on skewed data.The idea is to compare if preprocessing techniques work better when there is an o..

Nineth competition following Youhan Lee's curriculum. Regression competition using tabular data. Zillow Prize: Zillow’s Home Value Prediction (Zestimate)Can you improve the algorithm that changed the world of real estate?www.kaggle.comFirst Kernel: Simple Exploration Notebook - Zillow PrizeEDA kernel focused on univariate correlation analysis.Insight / Summary:1. Removing outliersulimit = np.per..

Eighth competition following Youhan Lee's curriculum. Regression competition using tabular data. New York City Taxi Trip DurationShare code and data to improve ride time predictionswww.kaggle.comFirst Kernel: Dynamics of New York city - AnimationUse K-means clustering to cluster New York into different groups based on location, and analyze the traffic into and out of every cluster as a function ..

Seventh competition following Youhan Lee's curriculum. Object segmentation(Deep learning) competition. 2018 Data Science BowlFind the nuclei in divergent images to advance medical discoverywww.kaggle.comFirst Kernel: Teaching notebook for total imaging newbies"This kernel will implement classical image techniques and will hopefully serve as a useful primer to people who have never worked with im..

Fifth Competition following Youhan Lee's curriculum. Multi-class classification competition using tabular data.  Costa Rican Household Poverty Level PredictionCan you identify which households have the highest need for social welfare assistance?www.kaggle.comFirst Kernel: A Complete Introduction and WalkthroughOverall walkthrough on the competition: from eda to modeling.Insight / Summary:1. Metr..

Sixth Competition following Youhan Lee's curriculum. Multi-class classification competition using image data. TensorFlow Speech Recognition ChallengeCan you build an algorithm that understands simple speech commands?www.kaggle.comFirst Kernel: Speech representation and data explorationMainly about exploring speech data.Insight / Summary:1. VisualizationThere are two theories of a human hearing -..

This post heavily relies on Will Koehrsen's public codes on Home Credit Default Risk Competition. https://www.kaggle.com/code/willkoehrsen/start-here-a-gentle-introductionhttps://www.kaggle.com/code/willkoehrsen/introduction-to-manual-feature-engineering  Above 2 posts are the first 2 public codes(kernels) on this competition written by Koehrsen and they are reviewed from my previous post:  [Kag..

Fourth Competition following Yuhan Lee's curriculum. Binary classification competition using image data. Statoil/C-CORE Iceberg Classifier ChallengeShip or iceberg, can you decide from space?www.kaggle.comFirst Kernel: Keras Model for Beginners (0.210 on LB)+EDA+R&DEDA + basic modelingUsing Keras which makes building deep neural networks very easy.Basic conv net model.Insights/Summary:1. Calcula..

Third Competition following Yuhan Lee's kaggle curriculum. Binary classification competition using tabular data. First KernelGeneral EDA kernel. Insights/Summary:1. This kernel's eda was processed by this order:target column(binary var) analysisMissing values examinationdata type checkingunique classes checking for object(categorical) columnsencoding categorical columnsanomalies detectioncheckin..

There are several types of correlation analysis:Pearson CorrelationMost commonly used methodMeasures linear relationshipsSuitable for continuous variablesValues range from -1 to 1Assumes normal distributionSpearman CorrelationMeasures rank-based correlationCaptures non-linear relationships wellSuitable for ordinal dataLess sensitive to outliersCan be used even when data is not normally distribut..

Skip Connection?Skip connections are now a standard module in many convolutional architectures.They provide alternative paths for gradients (including backpropagation). This additional pathway has been experimentally proven to often help with model convergence.In deep architectures, skip connections, as the name suggests, skip some layers of the neural network and provide the output of one layer..