Data now informs decisions in almost every organisation, from loan approvals and stock planning to disease surveillance and crop advisory services. Professionals who cannot work with it directly are left relying on others to interpret it, and those who manage data projects without understanding the basics struggle to tell a sound model from a misleading one. A solid grounding in how data is prepared, analysed and modelled has become a core workplace skill rather than a specialist one.

The Fundamentals of Data Science & Machine Learning in Practice Training Course assumes no programming background. Days one to five cover the data science workflow, Python essentials, data cleaning and exploration with pandas, descriptive statistics, visualisation and a first regression model. Days six to ten go further: classification and model evaluation, feature preparation and pipelines, ensemble methods, clustering and dimensionality reduction, plus tasters in time series, text analysis and neural networks. Responsible practice, including bias, data protection and clear communication of uncertainty, runs through the programme and is drawn together on the final day.

Teaching moves at a steady pace with live coding demonstrations, short practice tasks after every concept and plenty of support from the facilitator. Datasets are drawn from familiar settings such as household surveys, mobile money transactions, clinic records and retail sales. Participants finish by completing a small project end to end and presenting it, leaving with working notebooks they can reuse and a realistic plan for continuing to learn.

On completing the ten days, participants will be able to:

• Describe the data science workflow and where machine learning fits within it.
• Write basic Python code in Jupyter notebooks to load, inspect and manipulate data.
• Clean, filter, group and merge tabular data using pandas.
• Summarise data with descriptive statistics and present it in clear Matplotlib and Seaborn charts.
• Build regression and classification models with scikit-learn and evaluate them with suitable metrics.
• Apply train-test splits, cross-validation and simple pipelines to avoid overfitting and data leakage.
• Group records into meaningful segments using k-means clustering.
• Explain model results, limitations and ethical risks in plain language to non-specialist colleagues.

No coding experience is required. The programme suits newcomers to data work such as:

• Business and data analysts who currently rely on Excel
• Monitoring and evaluation officers in NGOs and development programmes
• Finance, credit and operations staff in banks, SACCOs and microfinance institutions
• Researchers, statisticians and graduate students moving into machine learning
• Health records and health information officers
• Government planning, statistics and policy officers
• IT and systems staff planning a move into data science
• Managers and project leads who commission or oversee data projects
• Marketing and customer insight staff in telecoms and retail

By the final session, each participant takes away:

• A working Python and Jupyter setup, plus a starter notebook template for new analyses.
• A pandas reference sheet covering the cleaning and reshaping steps used most often.
• A completed exploratory analysis with charts and written findings on a realistic dataset.
• Trained regression and classification models with their evaluation reports.
• A customer or beneficiary segmentation built with k-means clustering.
• A simple checklist for spotting overfitting, data leakage and biased training data.
• A finished end-to-end mini-project, presented to the group, to start a personal portfolio.
• A 90-day learning plan naming the next skills, datasets and resources to tackle.

The pace is set for people who are new to code, and every idea is practised straight away. The approach includes:

• Live coding demonstrations that participants follow line by line in their own notebooks
• Short practice tasks after each concept, with the facilitator on hand to troubleshoot
• Pre-built starter notebooks so that time goes on understanding rather than typing
• Realistic datasets from survey, banking, health and retail settings
• Small-group exercises that pair participants with different levels of confidence
• A guided capstone project built step by step and presented on the last day

A CPD-accredited Certificate of Completion issued by Vision Reach Global Consultancy is awarded to each participant who attends the full ten days and completes the daily lab exercises and the guided capstone project.

The certificate records the course title, dates and delivery format. Together with the notebooks and project produced in class, it documents a first step into data science for appraisals or further study.