Darko Medin’s Post

🍷 Excited to share my latest project using the Wine dataset! 📊 I've created a notebook comparing PCA-transformed data with normal data, analyzing results and visualizing covariance correlations using a Random Forest classifier. 🌟 🔍 In this project, I explored the impact of PCA on data representation and the effectiveness of Random Forest in classification tasks. The visualizations provided fascinating insights into the data's structure and relationships. 📈 Whether you're into machine learning or just love exploring datasets, I hope this notebook sparks some interest! Check out the Kaggle notebook https://lnkd.in/e-TcNN9d and let me know your thoughts. 🚀 #MachineLearning #DataScience #RandomForest #PCA #WineDataset #DataVisualization

Iris Flower Classification! 🌿I'm thrilled to announce the completion of my recent data science project where I built a model to classify Iris flowers into different species. 🌼 This project has been an enriching learning journey, empowering me to apply diverse data science techniques and tools effectively. Dive into the video below for an in-depth exploration of my project! 📊💡 📊Key highlights: - Data Preprocessing - Exploratory Data Analysis (EDA) - KNeighborsClassifier ,LogisticRegression & Support Vector Machine Building & Evaluation - Data Visualization & Correlations #CodSoft #DataScience #IrisFlower #Classification #MachineLearning

Task 3: Iris flower classification🌸 Completed! I recently finished a project in which I used the IRIS dataset to create a model that would categorize Iris flowers into species according to their petal length petal width, sepal width and sepal length characteristics. I cleaned and preprocessed the data, used several graphs for exploratory data analysis, and created new features to increase the accuracy of the model based on the conclusions I drew. I experimented with a number of machine learning models before choosing the top one. #IrisClassification #CodSoft #DataScience #PredictiveModeling

🚀 Excited to Share My Latest Achievement! 🚀 I’m thrilled to announce that I’ve successfully completed Task-2 in Prodigy for Data Science! 🎉 In this task, I worked extensively on data cleaning and exploratory data analysis using the Titanic dataset from Kaggle. Addressed missing values and handled outliers to improve data quality. Standardized data formats and corrected inconsistencies to ensure accuracy. #DataScience #DataCleaning #ExploratoryDataAnalysis #TitanicDataset #MachineLearning #CareerGrowth #DataAnalysis

I am thrilled to announce that I have completed my task 2 on iris flower classification. I have analysing the data, visualized the data using the library matplotlib library and modeled using the Logistic Regression. #intenship #codsoft #datascience

Day 11 of #100DaysOfMLChallenge : Today was all about diving deep into Exploratory Data Analysis (EDA)! Excited to share that I explored the power of univariate analysis, understanding individual variables, and bivariate analysis, unraveling relationships between two variables. Unraveling insights from a single lens with univariate analysis is like discovering hidden gems within each feature. On the other hand, bivariate analysis takes the exploration to a whole new level by unveiling connections and correlations between different variables. ------- Adding an extra layer to my toolkit, I delved into the world of Pandas Profiling! This tool is a game-changer for quick and comprehensive data profiling, helping to streamline the EDA process. Excited about the progress and the depth of understanding gained today! Ready for more challenges and learning on this exciting journey! #DataScience #MachineLearning #100DaysOfCode #DataAnalytics #PandasProfiling #DataExploration #LearningJourney

Steven Paul Sanderson II, MPH, the maintainer and creator of several R packages has released new updates for the TidyDensity R package. https://lnkd.in/dnzvisny If you are new to this package, TidyDensity is great for R simulations in Data Science and Statistics and has many simplified random number generator implementations which are essential for simulations. Further TidyDensity outputs the data as a tibble adapted for further use in the tidyverse framework and can also be used for distribution estimation. Some new features and that cought my eye : -generally in forms of having a function ending with _aic(), _estimate), _tbl() Inverse Weibull distribution Zero truncated negative binomial distribution Generalized Beta Generalized Pareto Inverse Pareto distribution estimation Paralogistic distribution functions Overall around 15 new functions. In my opinion this is a great contribution to the R simulation framework and great work here Steven. Another great detail is that all the simulations have a complex data visualization outputs which are important for anyone working in this field. You find find the full list of changes here : https://lnkd.in/d9N9Upp7 #datascience #r #rprogramming #stats #biostatistician #simulaiton

"Titanic Data Adventures: Navigating Missing Values with FFill and BFill Magic! 🚢✨ #DataScienceWonders #TitanicImputationJourney" 🔍 Data Imputation Delight: Filling the Gaps in Titanic Dataset Handling missing values is an art! Let's dive into the world of forward-fill (FFill) and backward-fill (BFill) techniques to make our Titanic dataset more robust. 🚢 Why FFill and BFill? FFill: Propagate the last known value forward. BFill: Propagate the next known value backward. 🌐 Impact on Data Integrity: Preserving the temporal order of data. Mitigating missing values without distorting the dataset. 🚀 Data Science Pro Tip: Choose the right imputation strategy based on the unique characteristics of your dataset! #DataScience #DataAnalysis #TitanicDataset #MissingValuesMagic 🎩📊 Here is An Example of Code

Embarking on Statistical Mastery in R: Proficient in Variance, SD, Covariance, Correlation, and Dynamic Plots. Currently tasked by Kolhandai Yesu to elevate my analytical skills. Excited for the journey ahead! #RStats #DataScience #Analytics 📊🚀

Hey LinkedIn crew! Day 16 of the 100 Days of Data Science Challenge brought a deep dive into feature selection using wrapper methods! 📊 Today's Learnings: - Explored wrapper methods for feature selection, understanding their role in optimizing model performance. - Applied exhaustive feature selection to the classic Iris dataset, leveraging the mlxtend library for a seamless process. - Discovered the power of mlxtend, making feature selection an easy and efficient task. 📈 Results Snapshot: Wrapper methods allowed for a fine-tuned selection of features, enhancing the model's efficiency. 🚀 What's Next: Tomorrow, venturing into more advanced techniques for refining models. #100DaysOfDataScience #FeatureSelection #WrapperMethods 📊 Check out the attached shortened version of my notes for insights! P.S. If you're on your own data journey, share your thoughts below! 🌱🤝