I will independently review your statistical or machine learning model
About this Gig
Can you trust your models results?
I provide an independent, scientifically grounded review of one existing statistical, forecasting, or machine learning model. As a physicist with experience in quantitative modelling and simulation, I examine assumptions, validation methods, performance metrics, uncertainty, and whether the conclusions follow from the available evidence.
You receive a concise PDF report identifying methodological risks and practical next steps.
BASIC: One focused methodological question; up to 8 pages of documentation.
STANDARD: One model, up to 20 pages and 200 relevant lines of code.
PREMIUM: One model, up to 35 pages and 400 relevant lines of code, with a prioritized improvement and validation plan.
This is a document-based review. I do not execute code, retrain models, process raw datasets, or certify model performance. Each order covers one model version and one evaluation setup. New questions or materials require a separate offer.
For unusual or complex cases, please message me before ordering.
FAQ
Which types of models can you review?
Statistical models, regression and classification models, conventional machine learning, and forecasting methods. Deep learning, LLM systems, custom simulation software, and highly specialized domains may require a custom offer. Please ask before purchasing if unsure.
Will you verify that my model works correctly?
I assess the methods and evidence you provide. I can identify risks and weaknesses in the validation approach, but this service does not include reproducing results, executing code, or certifying model performance.
What should I provide?
A short explanation of the model, its intended use, your main question, the training/validation approach, performance metrics, and relevant documentation. Standard and Premium may include selected code excerpts within the stated limits. Please provide one coherent submission at the start.
Can you review a complete code repository or dataset?
No. Standard and Premium cover only relevant, clearly identified code excerpts for static inspection. Running notebooks, configuring environments, cleaning data, and dataset auditing are not included.
Can I submit several models or versions?
Each package covers one model version and one evaluation setup. Comparing models, reviewing a pipeline with several models, or examining multiple datasets requires a custom offer.
What does one revision include?
One round of corrections to factual misunderstandings or clarification of existing findings, based on the original submission. New questions, files, analyses, or changes to the model are not included.
Do you guarantee a particular accuracy or business outcome?
No. The report is an independent, evidence-based professional assessment limited to the materials reviewed. It is not an audit certification, regulatory assessment, or guarantee of performance.
Can I send confidential data?
Please send only material you are entitled to share, preferably anonymized. Do not include personal or sensitive data, access credentials, or protected customer information. If disclosure restrictions apply, contact me before ordering.
Are meetings and implementation included?
o. The standard service is asynchronous, with written communication and a PDF deliverable. Calls, model fixes, implementation and training are separately scoped services, if offered.
What if my project exceeds the package limits?
Message me before ordering with a short scope summary. If the work is suitable, I can provide a separate custom offer through Fiverr specifying materials, outputs, price and timing.

