Python Trend Analysis & Modeling 1 hour meeting
Budget: ₹600 – ₹1,500 INR
I have a raw dataset sitting in a database that I need turned into clear, insight-ready material. The first step is a Python script that connects directly to the source, pulls the data, and cleans it—handling missing values, outliers, and any structural quirks. Once the data is tidy, I want a short write-up that spells out the final sample size and the key variables you kept or engineered so my team can reproduce the work.
From there, the focus shifts to trends. I’m particularly interested in consumer preference patterns over time, but I also want you to surface any other meaningful movements the data reveals. Use pandas with libraries such as seaborn, matplotlib, or Plotly to build a set of crisp, publication-quality visuals—line charts, heat maps, and whatever else best conveys the story.
Next, apply three supervised learning models—Logistic Regression, SVM, and Random Forest. I need performance metrics (accuracy, precision-recall, ROC, etc.) and visuals that make it easy to compare how each model captures those trends. Overlay predictions against actuals so the narrative is obvious to a non-technical reader.
Please finish with a concise insights summary I can drop straight into a report. Emphasize actionable takeaways that marketing and product teams will care about.
Deliverables:
• Python notebooks or scripts with clear comments
• Cleaned dataset (CSV or preferred format)
• Trend graphs and consumer preference charts (high-resolution PNG or interactive HTML)
• Model performance visuals and comparison plots
• One-page insights summary explaining sampling decisions, key variables, and what the trends mean for the business
From there, the focus shifts to trends. I’m particularly interested in consumer preference patterns over time, but I also want you to surface any other meaningful movements the data reveals. Use pandas with libraries such as seaborn, matplotlib, or Plotly to build a set of crisp, publication-quality visuals—line charts, heat maps, and whatever else best conveys the story.
Next, apply three supervised learning models—Logistic Regression, SVM, and Random Forest. I need performance metrics (accuracy, precision-recall, ROC, etc.) and visuals that make it easy to compare how each model captures those trends. Overlay predictions against actuals so the narrative is obvious to a non-technical reader.
Please finish with a concise insights summary I can drop straight into a report. Emphasize actionable takeaways that marketing and product teams will care about.
Deliverables:
• Python notebooks or scripts with clear comments
• Cleaned dataset (CSV or preferred format)
• Trend graphs and consumer preference charts (high-resolution PNG or interactive HTML)
• Model performance visuals and comparison plots
• One-page insights summary explaining sampling decisions, key variables, and what the trends mean for the business