Person

Tim Young

Professor | Forestry, Wildlife and Fisheries

Specialization: Data science applications in machine learning and AI

Overview

His current focus areas in the data sciences are: advanced analytical modeling of real-time processes; elementary AI applications; data fusion and Total quality Data Management (TqDM), statistical visualization; real-time predictive analytics, and real-time statistical process control (SPC). Specific applications in the data sciences are focused on real-time predictions of product attributes for sustainable biomaterials using a host of algorithms that process in parallel. Past research projects have resulted in successful prediction of tensile strength and modulus of sustainable biomaterials in validation within 10% RMSEP. Algorithms currently being studied are Bayesian Additive Regression Trees (BART), deep learning NN with generalized bounds, bootstrap forests and boosted trees. Variable preselection using genetic algorithms (GA) with deep learning NN is also under investigation. Real-time data fusion and TQDM are essential foundations to advanced analytical modeling of real-time processes. Studies are ongoing in the fusion of multiple networked data sources with varying record lengths and time-lagging. Imputation methods are also being studied as a critical component of TQDM. Statistical process control (SPC) applications include: 'control bands for data signatures and footprints'; multivariate control charts; and the development of autocorrelated control charts. He teaches highly successful industry courses in ‘Advanced Analytics and Data Mining,’ ‘Process Analytics, Statistical Process Control | Lean Methods,’ and ‘Design of Experiments.’

Research Focus

Data Science as applied to the 'SMART' manufacturing and agriculture. Concentrations are in Total quality Data Management (TqDM), machine learning and AI applied sciences. Real-time predictive analytics and statistical process control (SPC).

Research Questions
  • 1. The greedy algorithm partition problem, i.e., minimizing the generalized error of prediction using algorithms related to BART, bootstrap forests and boosted trees.
  • 2. Bridging the gap between theoretical data science research in machine learning and AI, to applied research and direct applications in SMART manufacturing and SMART agriculture
  • 3. Enhanced fusion of tree-based algorithms to improve design of experiments (DOE) for rapid innovation
  • 4. Real-time data fusion and data quality assessment and improvement to enhance Total quality Data Management (TqDM)
Courses
Below are courses taught during the current or past three academic years. Consult Timetable for the most current listing of courses and instructor(s).

FORS 590 - Advanced Topics in Forestry
1 - 3 credit hours

Recent advances and concepts; research techniques and analysis of current problems.

Repeatability: May be repeated. Maximum 6 hours.
Registration Permission: Consent of instructor.

Other Instructors: Taylor, Adam | Poudyal, Neelam Chandra | Jean-Philippe, Sharon

FORS 593 - Independent Study in Forestry
1 - 4 credit hours

Repeatability: May be repeated. Maximum 6 hours.

Other Instructors: Kojima, Cheryl Jean | Li, Mi | Cui, Shaoqing | Wang, Siqun

FWF 590 - Advanced Topics in Forestry, Wildlife and Fisheries
1 - 3 credit hours

Recent advances and concepts, research techniques, and analysis of current problems.

Repeatability: May be repeated. Maximum 6 hours.
Registration Permission: Consent of instructor.

Other Instructors: Kwit, Charles | Graves, Christopher Austin | Poudyal, Neelam Chandra | Hodges, Donald G

FWF 690 - Advanced Topics in Natural Resources
1 - 3 credit hours

Exposure and in-depth training in contemporary topics and approaches important to advanced research in natural resources.

Repeatability: May be repeated with consent of department. Maximum 9 hours.
Registration Restriction(s): Minimum student level – graduate.

Tim Young image
427 Plant Biotechnology Building
2505 E J Chapman Drive
Knoxville, TN 37996-4563
Education and Training
  • Doctorate, Statistics, General, Univ of Tennessee Knoxville*, 2007
  • MS, Statistics, Other, Univ of Tennessee Knoxville*, 1993
  • MS, Forestry, General, Univ of Wisconsin Madison*, 1983

Tim Young

Professor | Forestry, Wildlife and Fisheries
Tim Young image
427 Plant Biotechnology Building
2505 E J Chapman Drive
Knoxville, TN 37996-4563
Education and Training
  • Doctorate, Statistics, General, Univ of Tennessee Knoxville*, 2007
  • MS, Statistics, Other, Univ of Tennessee Knoxville*, 1993
  • MS, Forestry, General, Univ of Wisconsin Madison*, 1983
Overview

His current focus areas in the data sciences are: advanced analytical modeling of real-time processes; elementary AI applications; data fusion and Total quality Data Management (TqDM), statistical visualization; real-time predictive analytics, and real-time statistical process control (SPC). Specific applications in the data sciences are focused on real-time predictions of product attributes for sustainable biomaterials using a host of algorithms that process in parallel. Past research projects have resulted in successful prediction of tensile strength and modulus of sustainable biomaterials in validation within 10% RMSEP. Algorithms currently being studied are Bayesian Additive Regression Trees (BART), deep learning NN with generalized bounds, bootstrap forests and boosted trees. Variable preselection using genetic algorithms (GA) with deep learning NN is also under investigation. Real-time data fusion and TQDM are essential foundations to advanced analytical modeling of real-time processes. Studies are ongoing in the fusion of multiple networked data sources with varying record lengths and time-lagging. Imputation methods are also being studied as a critical component of TQDM. Statistical process control (SPC) applications include: 'control bands for data signatures and footprints'; multivariate control charts; and the development of autocorrelated control charts. He teaches highly successful industry courses in ‘Advanced Analytics and Data Mining,’ ‘Process Analytics, Statistical Process Control | Lean Methods,’ and ‘Design of Experiments.’

Research Focus

Data Science as applied to the 'SMART' manufacturing and agriculture. Concentrations are in Total quality Data Management (TqDM), machine learning and AI applied sciences. Real-time predictive analytics and statistical process control (SPC).

Research Questions
  • 1. The greedy algorithm partition problem, i.e., minimizing the generalized error of prediction using algorithms related to BART, bootstrap forests and boosted trees.
  • 2. Bridging the gap between theoretical data science research in machine learning and AI, to applied research and direct applications in SMART manufacturing and SMART agriculture
  • 3. Enhanced fusion of tree-based algorithms to improve design of experiments (DOE) for rapid innovation
  • 4. Real-time data fusion and data quality assessment and improvement to enhance Total quality Data Management (TqDM)
Courses
Below are courses taught during the current or past three academic years. Consult Timetable for the most current listing of courses and instructor(s).

FORS 590 - Advanced Topics in Forestry
1 - 3 credit hours

Recent advances and concepts; research techniques and analysis of current problems.

Repeatability: May be repeated. Maximum 6 hours.
Registration Permission: Consent of instructor.

Other Instructors: Taylor, Adam | Poudyal, Neelam Chandra | Jean-Philippe, Sharon

FORS 593 - Independent Study in Forestry
1 - 4 credit hours

Repeatability: May be repeated. Maximum 6 hours.

Other Instructors: Kojima, Cheryl Jean | Li, Mi | Cui, Shaoqing | Wang, Siqun

FWF 590 - Advanced Topics in Forestry, Wildlife and Fisheries
1 - 3 credit hours

Recent advances and concepts, research techniques, and analysis of current problems.

Repeatability: May be repeated. Maximum 6 hours.
Registration Permission: Consent of instructor.

Other Instructors: Kwit, Charles | Graves, Christopher Austin | Poudyal, Neelam Chandra | Hodges, Donald G

FWF 690 - Advanced Topics in Natural Resources
1 - 3 credit hours

Exposure and in-depth training in contemporary topics and approaches important to advanced research in natural resources.

Repeatability: May be repeated with consent of department. Maximum 9 hours.
Registration Restriction(s): Minimum student level – graduate.