Application of Machine Learning in Workforce Planning: Artificial Intelligence in Human Resource Management

This week I finalised my assignment, refined its formatting and converted it into the IEEE template. A big challenge I faced was having to rework all the references into the numbered by appearance configuration, which was eventually sorted out. After completing the report writing and review, I drafted and then timed a script for the seminar presentation. Initially, my draft script lasted 18 minutes as I wanted to ensure I spoke at a coherent and audible pace. I then removed a large portion of redundant information and reworded certain sections of the presentation script better to convey the inputs and outputs of the research. Eventually, the presentation reached below 11 minutes with a mildly-fast speaking pace.

Finally, the completed seminar was recorded and then uploaded to youtube. The final report document’s cover page includes a link to the seminar. Overall, I am happy with the research and results, and hopefully, the effort and interest put into the work will be recognised.

Following the previous week’s feature-engineering success, the dataset was ready for analysis and machine-learning model development. Scripting occurred to setup four models using the Linear Regression, Decision Tree Classification, Random Forest Classification and Support Vector Machine algorithms. Each was independently tested, and metrics were produced for each, including the models’ accuracy, precision, recall and aggregated F1-score. Visualisations were also produced.

Writing for the assignment is ongoing, and a first draft totalling over 4,500 words, has been completed. As such, subsequent revisions will take place to further refine and reduce the word count. Over the next week these revisions will take place followed by formatting to the IEEE format as well as beginning to create the script outline to be used for the seminar presentation.

This week, I continued my work identifying a suitable dataset for developing workforce planning machine learning models. Eventually, I decided to use the employee absenteeism dataset available on Kaggle.com as the base. The inherent challenge with the dataset, however, was its lack of a binary variable to use for classification models. Moreover, the dataset’s numeric variables, i.e., length of service and absent hours, would not provide enough depth to train and test any machine learning models adequately. As a result, significant effort was made to aggregate various variables against each record, such as the number of a specific job title per store or the turnover rate of a position, store location and estimated turnover likelihood per employee record. As a result, several variables were created based on existing data that was used to define a single classification for training and testing called “Workforce planning risk”.  This variable then helped create a secondary classification called “Recruitment required?” this would check if a record were terminated (1) and had a high risk (1) and then would classify the record as True. This variable will now be used to develop any machine learning models, which will be completed over the next week.

Since last week, I have begun structuring assignment four. Fortunately, feedback was received for assignment two (Project Proposal and Plan) midway through this week, which has significantly helped shape the direction of the final research project and its results. The feedback received encouraged using open-use datasets from sites such as Kaggle and AIHR to test the application of machine learning for workforce planning. I believe these datasets meet the minimum requirements to develop a basic machine-learning model to test various workforce planning variables. However, I do expect to encounter issues with the potential lack of demand and supply drivers from these datasets to help influence the machine learning models. For example, without knowing the required staffing rates for certain roles or other business-specific factors such as customer-to-staff ratios, the model will effectively assess the intersection between recruitment and turnover as a proxy for “Workforce planning”. Additional literature reviews will be conducted to source content discussing workforce planning drivers and detractors, which will be included in the final report.

Over the next week, I will finalise sourcing datasets to train and develop and machine-learning model. Then I will begin structuring the final results in the assignment four report.

This week I finished the draft annotated bibliography, reviewed and refined it, and then submitted it via easts. Overall, the annotated bibliography investigation and writing task was extremely interesting. The structured approach to analysing and synthesising the most important elements of the reviewed articles helped me grasp their concepts and findings better. However, I struggled to remain within the word limit for several articles as so much content could be discussed. As such, several annotated bibliography sections had to be reworked and reduced to fit the 2,250 word limit. A reflection was also written for assignment three which noted several discoveries I made relating to my research topic and described how these findings will influence the future direction of my project’s final report.

Next week, I will begin working through Assignment four. I have completed the majority of work for my other subjects and which will allow me to focus on the impending due date for this subject.

This week I continued working through drafting the annotated bibliography sections and have so far written paragraphs on  7 of the minimum 12 required articles. Since the final project will be an extended bibliography, I am trying to conserve more a more extensive review for that assignment. However, I have found several linkages between the reviewed articles and highlighted as such within the bibliography. At this pace, I expect to complete a first draft by the end of next week as long as my other assignments are completed on time as well. No issues to note from this week of study and I will continue sourcing additional articles as required.

I spent time reviewing literature relating to machine learning’s application in various business functions, the implications of artificial intelligence and machine learning within human resource management, workforce planning methodologies, and recent innovations in these areas.
Interestingly, I found a gap in the literature on machine learning in human resource management that did not relate to turnover and recruitment. I needed to look into other areas, such as performance management and learning systems, to find researchers noting opportunities for further study. I believe these articles help support the justification for this project in investigating other applications of machine learning within human resource management.

Since last week, assignment two has been submitted, and work has now begun on drafting sections of the annotated bibliography. I have other assignments due in the upcoming fortnight, which I will separate my time across, but I plan to utilise the extended easter holiday period to make progress on assignments three and four.

This week the draft for the project proposal and plan assignment has been completed. I received feedback for the first assignment, which recommended refining the research questions and scope. As such, revisions were made within the project proposal and plan to centralise its’ objectives around four related questions instead of five. The upcoming week will be used to review the draft and then resubmit in the easts portal.

For the upcoming week’s work, I have structured the annotated bibliography literature search into four distinct categories to help differentiate each literature type which should support future discussions of results and linking related literature together.

Unfortunately, progress this week was impacted by other commitments. I was able to spend some time sourcing additional literature for review, which I will begin to analyse and synthesise for assignment three. I made moderate progress on the work breakdown structure, Gantt chart and risk analysis this week with the intention to have them completed by Sunday. I believe that I am still on time with my progress and, despite some delays, will have completed assignment two at least one week prior to the due date.