drclab commited on
Commit
19ed75c
1 Parent(s): 419f607
Files changed (4) hide show
  1. CV2023.pdf +0 -0
  2. cv.txt +158 -0
  3. pdf.ipynb +0 -0
  4. prompt_cv.ipynb +120 -0
CV2023.pdf ADDED
Binary file (225 kB). View file
 
cv.txt ADDED
@@ -0,0 +1,158 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ C.J. DUAN
2
+ Data Scientist. Ph.D
3
+ @Data.Scientist@dulun.com ♂phone757-742-3896 ♂¶ap-¶arkerDRC Lab, Virginia /gl⌢bewww.dulun.com
4
+ /linkedincj-duan-3ab27a191 /githubhublun
5
+ EXPERIENCE
6
+ Data Scientist (contract)
7
+ Fortune 500 Company (CPG), Research Data Science
8
+ Ὄ52021 - 7.2022 ♂¶ap-¶arkerRemote
9
+ •Input Data (Snowflake on AWS): Time-series of weekly sales data and
10
+ weekly media (advertising) spend and GRP (Impression)s
11
+ •Model: State-Space Bayesian Media Mix Model using R and RStan
12
+ •Output Results: ROI effectiveness in terms of MC (media contribution)
13
+ over MS (media spend)
14
+ Chief Research Data Scientist
15
+ DRC Lab @ Dulun Research & Consulting
16
+ Ὄ52016 - Now ♂¶ap-¶arkerFreelance
17
+ Completed Projects :
18
+ •(2020-2021) Algorithmic Trading Robot, coded in Python, digests recent
19
+ (commodity future) prices and execute order composition and routing
20
+ automatically.
21
+ FFT, Risk Control, Trading
22
+ •(2016-2017) Develop the best distribution routing model, which saved
23
+ an Alabama flower-growing farmer thousands fuel dollars on monthly
24
+ basis.
25
+ Transportation Scheduling, Constraint Programming, OR
26
+ Replication & Learning Projects (2016 - now) :
27
+ •Analysis of 6.4 million SARS-CoV-2 genomes identifies mutations
28
+ associated with fitness (Obermeyer, et. al. , 2022, Science )
29
+ Bayesian multinomial logistic regression model using Pyro
30
+ •Survival analysis in Stan (Filipe S. Dias, 2022, Web Blog )
31
+ Survival Analysis
32
+ •Deep learning recommendation model for personalization and
33
+ recommendation systems (Naumov, et. al. 2019, ArXiv )
34
+ TorchRec
35
+ •Bayesian Inference for a Generative Model of Transcriptome Pro-
36
+ files from Single-cell RNA Sequencing (Lopez, et. al. , 2018, Na-
37
+ ture Methods )
38
+ Variational Inference for Generative Model of scRNA seq using PyTorch and Pyro
39
+ •Bayesian aggregation of average data: An application in drug de-
40
+ velopment (Weber, et. al. , 2018, Annals of Applied Statistics )
41
+ Pharmacometric Model using RStan
42
+ Assistant Professor of Quantitative Methods
43
+ Troy University
44
+ Ὄ5March 2009 – May 2017 ♂¶ap-¶arkerTroy, Al
45
+ Research Projects :
46
+ •Research Paper (2021): Team Contingent or Sport Native? A Bayesian
47
+ Analysis of Home Field Advantage,
48
+ •Research Paper (2021): Exposing model bias in machine learning revisit-
49
+ ing the boy who cried wolf in the context of phishing detectionWORK PHILOSOPHY
50
+ “Modeling in the sciences is in fact al-
51
+ most always generative modeling”
52
+ Kingma and Welling (2019)
53
+ MOST PROUD OF
54
+ /gavelHidden Markov Model (HMM)
55
+ state space, transition, and emission
56
+ /binocularsMarkov Decision Process (MDP)
57
+ states, actions, rewards
58
+ /cogsPyro & PyTorch
59
+ SVI (stochastic variaitonal inference) based
60
+ Variational Autoencoder
61
+ ⌛Multi-Arm Bandit (MAB) and RL
62
+ 𝜋policy mapping actions to contexts
63
+ /meetupMedia Mix Modeling
64
+ ROAS, mROAS, Adstock, Hill function
65
+ ἲ1A/B Testing and MCI
66
+ multiple causal (treatment) inference
67
+ ὖELDA / Mixture / Clustering
68
+ Latent Dirichlet Allocation, Topic Models
69
+ Ὠ0Stan & Bayesian Modeling
70
+ data, parameters, model, and prediction
71
+ Ἴ6Servire Est Vivere
72
+ helping Braydon Farm LLC in Alabama opti-
73
+ mize their daily flower delivery routes using
74
+ Transportation Model
75
+ EDUCATION
76
+ Ph.D in Industrial Management
77
+ Clemson University
78
+ Ὄ52000 – 2007
79
+ REMARKS
80
+ "...On behalf of my wife and I, we would like to sin-
81
+ cerely thank Dr. Duan, Mrs. Amy Hu, and you for the
82
+ time you spent developing a distribution plan for our
83
+ new floral operation. We have incorporated it into our
84
+ operation and it appears to have significantly reduced
85
+ our travel time and expenses in distribution our prod-
86
+ ucts to the Marvin Stores chain. Hopefully this will bePUBLICATIONS
87
+ /file-altJournal Articles
88
+ •(Chaojie), D. C. J., & Chakravarty, A. (2021). Team contingent
89
+ or sport native? a bayesian analysis of home field advantage in
90
+ professional soccer. Journal of Business Analytics ,4(1), 67–75.
91
+ doi:10.1080/2573234X.2020.1854625
92
+ •(Chaojie), D. C., & Gaurav, A. (2021). Exposing model bias in ma-
93
+ chine learning revisiting the boy who cried wolf in the context
94
+ of phishing detection. Journal of Business Analytics ,4(2), 171–
95
+ 178. doi:10.1080/2573234X.2021.1934128. eprint: https:
96
+ / /doi.org/10.1080/2573234X.2021.1934128
97
+ •Duan, C. [C.J.], & Pierce, G. S. (2017). Starting operations man-
98
+ agement with a ‘Big Bang’: Using sitcom to introduce OM con-
99
+ cepts to students. Operations Management Education Review ,11,
100
+ 115–122.
101
+ •Duan, C. [CJ], Hu, J., Garrott, S., et al. (2016). Using excel solver
102
+ to solve Braydon Farms’ truck routing problem: A case study.
103
+ South Asian Journal of Management Sciences (SAJMS) ,10(1), 38–
104
+ 47. doi:10.21621/sajms.2016101.04
105
+ •Duan, C. [Chaojie], Grover, V., Roberts, N., & Balakrishnan, N. (
106
+ (2014). Firm valuation effects of the decision to adopt relation-
107
+ ally governed business process outsourcing arrangements. In-
108
+ ternational Journal of Production Research ,52(15), 4673–4694.
109
+ doi:10.1080/00207543.2014.884289. eprint: https:/ /doi.org/
110
+ 10.1080/00207543.2014.884289
111
+ •Kronenburg, M. A., Shetterly, D. R., Duan, C., Krishnamoorthy,
112
+ A., & Loutzenhiser, K. (2013c). The impact of contract design
113
+ on contract performance satisfaction. International Journal of
114
+ Management Research and Reviews ,3(4), 2673.
115
+ •Duan, C. [Chaojie], Grover, V., & Balakrishnan, N. ( (2009). Busi-
116
+ ness process outsourcing: An event study on the nature of pro-
117
+ cesses and firm valuation. European Journal of Information Sys-
118
+ tems ,18(5), 442–457. doi:10.1057/ejis.2009.38
119
+ /usersConference Proceedings
120
+ •Duan, C. [C.J.]. (2013). Retrieving hypothesis parameters from
121
+ real life events. In The 43rdsoutheast decison sciences institute
122
+ conference proceedings . SE DSI.
123
+ •Kronenburg, M. A., Shetterly, D. R., Duan, C., Krishnamoorthy,
124
+ A., & Loutzenhiser, K. (2013a). Public management skills needed
125
+ in contract development. In The southeastern conference for pub-
126
+ lic administration . SECoPA.
127
+ •Kronenburg, M. A., Shetterly, D. R., Duan, C., Krishnamoorthy,
128
+ A., & Loutzenhiser, K. (2013b). Public management skills needed
129
+ in contract development. In The southeastern conference for pub-
130
+ lic administration . SECoPA.
131
+ •Kronenburg, M. A., Shetterly, D. R., Duan, C., Krishnamoorthy,
132
+ A., & Loutzenhiser, K. (2012). The impact of contract design on
133
+ contractor performance - a second look. In 5thinternational pub-
134
+ lic procurement conference . IPPC.
135
+ •Duan, C. [C.J.]. (2004). Teams in TQM: An knowledge based
136
+ view. In The 35thdsi.
137
+ •Duan, C. [C.J.]. (2001). One machine scheduling of jobs to min-
138
+ imize total weighted tardiness. In The 37thsoutheast informs . SE
139
+ INFORMS.an on-going study and we will be able to “tweak” or al-
140
+ ter the system based on your recommendations as our
141
+ business grows..."
142
+ Stephen (Brad) Bradshaw Garrott
143
+ Owner, Braydon Farms, 2015
144
+ "...I work on projects focused on educating high school
145
+ students in Mississippi in entrepreneurship opportu-
146
+ nities and what resources are available to start-ups. I
147
+ use Excel frequently and I am thankful for being familiar
148
+ with excel formula’s like vlookup and IF. ... I wish there
149
+ was an Operations Management class because that
150
+ was my favorite in undergrad and where I find myself
151
+ passionate. I am considering entering the workforce
152
+ aiming for an Operations Management position based
153
+ on my success in your class and how much I enjoyed it.
154
+ I think your class is very important and I’m glad I had
155
+ the opportunity to learn from you. I am also open to
156
+ entrepreneurship opportunities as I am so active in the
157
+ entrepreneurship ..."
158
+ Russell, QM 3345, 2016
pdf.ipynb ADDED
File without changes
prompt_cv.ipynb ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cells": [
3
+ {
4
+ "attachments": {},
5
+ "cell_type": "markdown",
6
+ "metadata": {},
7
+ "source": [
8
+ "## [ChatGPT Prompt Engineering for Developers](https://learn.deeplearning.ai/chatgpt-prompt-eng/)"
9
+ ]
10
+ },
11
+ {
12
+ "cell_type": "code",
13
+ "execution_count": 2,
14
+ "metadata": {},
15
+ "outputs": [],
16
+ "source": [
17
+ "from auth import API_KEY\n",
18
+ "import openai"
19
+ ]
20
+ },
21
+ {
22
+ "cell_type": "code",
23
+ "execution_count": 3,
24
+ "metadata": {},
25
+ "outputs": [],
26
+ "source": [
27
+ "openai.api_key = API_KEY"
28
+ ]
29
+ },
30
+ {
31
+ "cell_type": "code",
32
+ "execution_count": 8,
33
+ "metadata": {},
34
+ "outputs": [],
35
+ "source": [
36
+ "def get_completion(prompt, model='gpt-3.5-turbo'):\n",
37
+ " messages = [{'role':'user', 'content': prompt}]\n",
38
+ " response = openai.ChatCompletion.create(\n",
39
+ " model=model,\n",
40
+ " messages = messages,\n",
41
+ " temperature = 0, # this is the degree of randomness of the model's output\n",
42
+ " )\n",
43
+ " return response.choices[0].message['content']"
44
+ ]
45
+ },
46
+ {
47
+ "cell_type": "code",
48
+ "execution_count": 5,
49
+ "metadata": {},
50
+ "outputs": [],
51
+ "source": [
52
+ "text = f\"\"\"\n",
53
+ "You should express what you want a model to do by \\ \n",
54
+ "providing instructions that are as clear and \\ \n",
55
+ "specific as you can possibly make them. \\ \n",
56
+ "This will guide the model towards the desired output, \\ \n",
57
+ "and reduce the chances of receiving irrelevant \\ \n",
58
+ "or incorrect responses. Don't confuse writing a \\ \n",
59
+ "clear prompt with writing a short prompt. \\ \n",
60
+ "In many cases, longer prompts provide more clarity \\ \n",
61
+ "and context for the model, which can lead to \\ \n",
62
+ "more detailed and relevant outputs.\n",
63
+ "\"\"\""
64
+ ]
65
+ },
66
+ {
67
+ "cell_type": "code",
68
+ "execution_count": 6,
69
+ "metadata": {},
70
+ "outputs": [],
71
+ "source": [
72
+ "prompt = f\"\"\"\n",
73
+ "Summarize the text delimited by triple backticks \\ \n",
74
+ "into a single sentence.\n",
75
+ "```{text}```\n",
76
+ "\"\"\""
77
+ ]
78
+ },
79
+ {
80
+ "cell_type": "code",
81
+ "execution_count": 9,
82
+ "metadata": {},
83
+ "outputs": [
84
+ {
85
+ "name": "stdout",
86
+ "output_type": "stream",
87
+ "text": [
88
+ "Clear and specific instructions should be provided to guide a model towards the desired output, and longer prompts can provide more clarity and context for the model, leading to more detailed and relevant outputs.\n"
89
+ ]
90
+ }
91
+ ],
92
+ "source": [
93
+ "response = get_completion(prompt)\n",
94
+ "print(response)"
95
+ ]
96
+ }
97
+ ],
98
+ "metadata": {
99
+ "kernelspec": {
100
+ "display_name": "Python 3",
101
+ "language": "python",
102
+ "name": "python3"
103
+ },
104
+ "language_info": {
105
+ "codemirror_mode": {
106
+ "name": "ipython",
107
+ "version": 3
108
+ },
109
+ "file_extension": ".py",
110
+ "mimetype": "text/x-python",
111
+ "name": "python",
112
+ "nbconvert_exporter": "python",
113
+ "pygments_lexer": "ipython3",
114
+ "version": "3.10.6"
115
+ },
116
+ "orig_nbformat": 4
117
+ },
118
+ "nbformat": 4,
119
+ "nbformat_minor": 2
120
+ }