RMSE vs MAE: Choosing the Right Metric for Model Evaluation

I expected Random Forest to win. It didn't. Built a sales forecasting model on ecommerce data. Tried both Linear Regression and Random Forest. Linear Regression got a lower RMSE. Random Forest overfit. That was a good reminder more complex doesn't always mean better. Sometimes the data is just... linear. The project also taught me that feature engineering matters more than model choice. Getting the right features in lag variables, rolling averages, trend components made a bigger difference than switching algorithms. GitHub link in the comments 👇 #MachineLearning #Python #SalesForecasting #DataScience #OpenToWork

Day 19/75 — One simple trick I use to understand any dataset 👇 Whenever I open a new dataset, I don’t start with complex analysis. I start with this: df.describe() 💡 What it gives you instantly: • Mean (average) • Min & Max values • Standard deviation • Count of values 👉 In just one line, you get a quick summary of your data. But here’s the important part: I don’t just look at the numbers. I ask: • Does this make sense? • Are there outliers? • Is something unusual? 🚨 Lesson: Before building anything… 👉 Understand your data first. This one habit saves me a lot of time. What’s the first thing you do when you open a dataset? 👇 #DataScience #Python #Pandas #LearningInPublic #OpenToWork

Some problems are not about searching everywhere, but about searching smartly. Day 26/100 — Data Structures & Algorithms Journey Today’s Problem: Two Sum II (Input Array is Sorted) This problem helped me understand how sorting can simplify searching and improve efficiency. Approach: Since the array is already sorted, I used the Two Pointer technique instead of brute force. I placed one pointer at the beginning and another at the end of the array. By adjusting pointers based on the sum: If the sum is too small → move left pointer If the sum is too large → move right pointer If equal → solution found This approach avoids unnecessary checks and works efficiently in a single pass. Key Takeaways: Sorted data can unlock optimized solutions Two Pointer technique reduces complexity significantly Thinking smart is better than brute force Efficient algorithms improve performance and clarity This problem strengthened my understanding of pointer techniques and optimization strategies. #DSA #LeetCode #TwoPointers #ProblemSolving #SoftwareEngineering #CodingJourney #100DaysOfCode #TechLearning #DeveloperJourney #Programming #Python #InterviewPreparation #CodingSkills #ComputerScience #FutureEngineer #TechCareers #SoftwareDeveloper #LearnInPublic #OpenToWork

🚀 Day 34/70 – Correlation in Statistics Today I learned about Correlation 📊 Correlation measures the relationship between two variables. 📌 Types of Correlation 1️⃣ Positive Correlation Both variables increase together 2️⃣ Negative Correlation One increases while the other decreases 3️⃣ No Correlation No relationship between variables 📌 Correlation Coefficient (r) Value ranges from -1 to +1 ✔ +1 → Perfect positive correlation ✔ -1 → Perfect negative correlation ✔ 0 → No correlation 📌 Python Example import numpy as np x = [10, 20, 30, 40, 50] y = [15, 25, 35, 45, 55] correlation = np.corrcoef(x, y) print(correlation) 📊 Why Correlation is Important ✔ Identifies relationships in data ✔ Helps in prediction ✔ Used in feature selection ✔ Important for machine learning Today’s Learning: Correlation helps understand how variables are connected 🔥 Day 34 completed 💪 Becoming more analytical every day! #Day34 #Statistics #DataAnalytics #Python #LearningInPublic #FutureDataAnalyst #70DaysChallenge

🚀 Why do customers leave a company? I recently worked on a Customer Churn Prediction Project to find out—and the results were surprising. 🔧 Tech Stack: Python | Pandas | NumPy | Scikit-learn | Matplotlib 📊 What I did: Cleaned and analyzed customer data Built ML models (Logistic Regression, KNN) Tuned hyperparameters using GridSearchCV 💡 Key Insight: Customers with month-to-month contracts were significantly more likely to churn compared to long-term contract users. 📈 The model achieved ~85% accuracy in predicting churn. 🔗 I’ve shared the full project on GitHub (link in comments). Would love your feedback! 🙌 #MachineLearning #DataScience #Python #Projects #OpenToWork

#Matrix in NumPy Arrays – Quick Concept In NumPy, a matrix is represented using a 2D array (rows × columns) and is widely used in data analysis and analytics workflows. 🔹 Creating a Matrix #Python import numpy as np A = np.array([[1, 2, 3], [4, 5, 6]]) 🔹 Common Operations A + A → Addition A * 2 → Scalar multiplication A @ B → Matrix multiplication A.T → Transpose 🔹 Special Matrices np.zeros() → Zero matrix np.ones() → Ones matrix np.eye() → Identity matrix 🔹 Why it matters Used in KPI calculations, data transformation, and machine learning models. #Key takeaway: NumPy arrays are faster and more flexible for matrix operations. #NumPy #Python #DataAnalytics #DataAnalyst #MIS #Analytics #OpenToWork

📊 NIFTY Market Analysis Dashboard I built this dashboard using Python to analyze market trends. 🔹 Visualized candlestick chart with Moving Averages (MA50 & MA100)   🔹 Added RSI indicator to understand momentum   🔹 Processed real market data using Pandas  🔸 Strategy Logic: - Buy condition: RSI < 35, MA50 > MA100 and price above MA50   - Sell condition: RSI > 65, MA50 < MA100 and price below MA50  This project helped me understand how indicators behave in real market conditions. 📌 Learning:   Indicators alone do not guarantee profit, but they help in better decision making. Tools Used: Python | Pandas | Plotly | Streamlit #Python #DataAnalysis #StockMarket #Nifty #Learning

I’m excited to share a new project I’ve been working on: an Iris Flower Classification system built using Python and Scikit-learn! 🧠 This project focuses on the fundamentals of Supervised Learning. By training a model on sepal and petal measurements, I was able to successfully classify three different species of Iris flowers with high precision. Key Highlights: Algorithm: Implemented K-Nearest Neighbors (KNN) for classification. Preprocessing: Used StandardScaler for feature scaling and split the data for robust testing. Performance: Achieved an accuracy of 95% - 100% on the test set. 📈 Working on this reinforced my understanding of the machine learning workflow from data loading and preprocessing to model evaluation using classification reports and confusion matrices.CodeAlpha You can check out the full code and README on my GitHub here: [ https://lnkd.in/gJcYm7dF ] 🚀 #MachineLearning #DataScience #Python #ScikitLearn #AI #CodingJourney #GitHub #Classification #CodeAlpha

Most people think data analysis starts with tools. Excel. Python. SPSS. Machine Learning. It doesn’t. It starts with a question. And this is where many get it wrong. Because a weak question will always produce weak insights, no matter how advanced your analysis is. I’ve seen projects where everything looked “technically correct”… but the conclusions made no real sense. Not because the data was bad. But because the question behind the analysis was shallow. Good analysis is not about running models. It’s about thinking clearly before you touch the data. So before your next project, ask yourself: Are you asking a question that actually matters… or just one that is easy to analyze? #HPAnalytics #DataAnalysis #Research #MachineLearning #CriticalThinking

All-in-One NumPy Guide. This guide includes: • Introduction to NumPy & array fundamentals • Indexing, slicing, and reshaping • Broadcasting and vectorization • Mathematical & statistical operations • And many more.... This “all-in-one” resource is designed to build a strong foundation in numerical computing and help apply NumPy effectively in Data Science and Machine Learning projects. """I’d really appreciate your feedback and would love to connect with professionals in the data field!""" #NumPy #Python #DataScience #MachineLearning #DataAnalytics #OpenToWork #Portfolio