A data-driven approach to customer lifetime value prediction using probability and machine learning models
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Customer lifetime value is an important marketing metric and has applications in market segmentation, strategy development, and direct marketing programs, especially when customers are not under contract. In this research, we demonstrate the prediction of the lifetime value of patients in a health service portfolio in two separate ways. The probability of a patient being alive and their value in the coming evaluation period are first predicted using a probability model that has been well-established in the marketing community. We then use several machine learning algorithms to perform the same task. The results of these two approaches are compared in terms of accuracy to gain insight into their respective strengths and weaknesses. We believe that the work is one of the first attempts to gain an understanding of the use of machine learning algorithms in this important marketing issue. The results showed that the probability model performs better than the machine learning models, probably due to the assumption required in the probability calculations. It is therefore recommended that an essential step in applying these software approaches is to verify the validity of the key assumption of regularity. In addition, in future studies, consideration should be given to a larger dataset with demographic variables beyond age and gender that were used in this study. Developing specific ML models for dealing with zero-inflated data, which is an inherent feature of customer lifetime data, will also be helpful. • Compare probability models and machine learning for predicting customer lifetime value. • Apply predictive analytics to assess patient value in a health service business • Evaluate model accuracy to understand strengths and weaknesses in decision-making. • Highlight the role of probability assumptions in customer lifetime predictions. • Explore the potential of machine learning for improving customer value estimation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it