Rural Micro Credit Assessment using Machine Learning in a Peruvian microfinance institution

Abstract

Microcredits are an important component in the development of the peruvian rural economy, which are granted by microfinance institutions, the assessment process for the rural an poor people has a high risk index which is traditionally controlled by the business rural advisor, whose main tasks are the evaluation and verification of the clients requesting these microcredits. This research proposes a model that presents the best level of assertiveness for microcredits assessment process based on determination analysis of rural variables based on the specialized literature in the area. This model serves as a decision-support tool for business rural advisor in order to reduce the credit risk of the rural microfinance institution, The most representative variables of the financial and microfinance segment have been evaluated; the data has been pre-processed; Machine Learning models have been selected, trained, validated and evaluated through different metrics. The most assertive model in the assessment process of the granting of rural microcredit, based on the variables and data used by the analyzed entity, are: the Artificial Neural Network (93.72), on Logistic Regression (86.07), Random Forest (66.35), Support Vector Machine (84.44), Decision Tree (88.80) and k-Nearest Neighbor (65.98). Finally, the level of assertiveness achieved by ANN model 93.72% is better than the entity traditional methodology 76.81%, showing an improvement of 16.91% in the index of default customers.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 18 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.