Skip to content
View ppblanco's full-sized avatar

Block or report ppblanco

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ppblanco/README.md

Pedro Pérez Blanco

Marketing and data analyst. MSc in Business Analytics (MUBA), Universidad Pontificia Comillas — ICADE, and a Marketing degree from ESIC. I work at the point where the two meet: business questions answered with models and data.

What I work on

  • Machine learning — supervised classification and regression, model comparison, interpretability with SHAP
  • NLP — topic modeling and text mining on multilingual corpora
  • Deep learning — recurrent architectures for time series forecasting
  • Data analytics — exploratory analysis, data quality, visualisation
  • Marketing analytics — lead scoring, audience and campaign analysis

Portfolio

pedro-perez-data.netlify.app

Six case studies, each with the method, the metrics and the code behind it.

Selected projects

Repository What it is
admissions-lead-scoring-rag Master's thesis. Lead scoring for university admissions: A/B/C prioritisation, temporal validation across intakes, and a RAG assistant. No data published.
retail-demand-forecasting-rnn Demand forecasting with SimpleRNN, LSTM, GRU and BiLSTM against a moving-average baseline, under strict chronological splits.
video-game-sales-prediction Supervised regression: baseline, linear, KNN and Gradient Boosting, with GridSearchCV, SHAP and metrics on two scales.
letterboxd-topic-modeling LDA on multilingual film reviews, and what it means when three of six topics turn out to be languages rather than themes.
atp-tennis-betting-analysis EDA of 11,794 ATP matches against betting odds, and why no simple strategy beats the house edge.
lipton-summer-shake-up Brand strategy and campaign for Lipton Ice Tea and Gen Z: mixed-method research, creative concept and budget.

Three of these are team projects; authorship is stated in each repository.

Contact

LinkedIn · Madrid, Spain

Popular repositories Loading

  1. video-game-sales-prediction video-game-sales-prediction Public

    Supervised regression to predict video game sales: baseline, linear, KNN and Gradient Boosting, with GridSearchCV, SHAP and metrics on two scales.

    Jupyter Notebook

  2. atp-tennis-betting-analysis atp-tennis-betting-analysis Public

    EDA of 11,794 ATP matches (2020-2024): player height, handedness and age against betting odds, and why no simple strategy beats the house edge.

    Jupyter Notebook

  3. letterboxd-topic-modeling letterboxd-topic-modeling Public

    LDA topic modeling on multilingual Letterboxd reviews: when 3 of 6 topics are languages, not themes. Reproducible NLP pipeline with spaCy and gensim.

    Jupyter Notebook

  4. retail-demand-forecasting-rnn retail-demand-forecasting-rnn Public

    Retail demand forecasting with RNNs: SimpleRNN, LSTM, GRU and BiLSTM against a moving-average baseline, under strict chronological train/val/test.

    Jupyter Notebook

  5. admissions-lead-scoring-rag admissions-lead-scoring-rag Public

    Machine learning lead scoring for admissions: A/B/C prioritisation, temporal validation across intakes and a RAG assistant. No data published.

    Jupyter Notebook

  6. lipton-summer-shake-up lipton-summer-shake-up Public

    Brand strategy campaign for Lipton Ice Tea and Gen Z: mixed-method research, creative concept and a merchandising budget that pays for itself.