Underscoring the Prospects of Machine Learning-based Valuation
in the Management of Public Assets in Nigeria.
ESV Abiodun Bewaji
The provision of effective management to publicly owned assets have been considered a crucial component of responsible governance across various countries of the world seeing that it entails the judicious utilization and maintenance of public resources in a manner that ultimately allows for improved benefit to the society. Assessing the monetary or economic worth of public properties, such as land, infrastructure and other tangible and intangible assets requires a clear understanding of value at a specific point in time. Asset valuation which stands out as a key tool in the management of public asset aids managers and policymakers in the public sector to determine the true market value of assets.
Asset valuation as an instrument for making informed policy decisions in terms urban planning, revenue generation, risk management and asset life cycle management usually have defined steps or procedures to be taken. The process of valuing public assets for purposes which includes financial reporting, budgeting and auditing to estimate returns or liabilities usually require large amount of data as a result making asset valuation time consuming and error prone amongst other difficulties faced under the traditional approach. Property valuation as growing field which have continued to gain wide recognition in the economic and business world have been evolving rapidly due to innovation that is driven by advancement in technology, data analytics and artificial intelligence.
The presence of innovative tools such as machine learning algorithms which brings greater accuracy, speed and objectivity to the process of valuation when predicting property values have complemented or replaced the traditional property valuation methods. The use of machine learning (ML) technique as a method of valuation to estimate the value of an asset, business or property, applies advanced algorithms to large datasets which enables more nuanced and data-driven valuation compared to the traditional methods such as comparative market analysis (CMA) which depends on historical data and static models.
In the face of emerging digital technologies in real estate, developing countries tends to have lagged behind in terms of machine learning implementation. Several challenges stemming from the scarcity of reliable public data, weak regulatory framework, poor digital infrastructure and limited expertise due to lack of education and training in machine learning have affected the application of enhanced valuation techniques in public asset management. Nigeria as a developing country which ranks amongst the largest economies in Africa is yet to grow its tech ecosystem and position itself as a leader in AI-driven valuation and data science in Africa.
It has become expedient that investment in data collection, digital infrastructure, education and regulatory framework should form the main focus and priority of government ministries, department and agencies operating in the real estate sector. Incorporating AI-enabled methods such as Regression Trees and Random Forest, Gradient Boosting Machines (GBMs) and Geospatial Analysis to public asset valuation would enhance the prediction of property values in the shortest possible time while considering input features including neighborhood quality, infrastructure proximity and trends in historical price. Leveraging the power of data-driven algorithms to improve the process of valuation across various sectors of the economy such as housing, energy and transport amongst others will offer significant advantages in terms of scalability, accuracy and adaptability.
The state of the Nigerian economy could be improved through the integration of disruptive digital technology which assesses public assets based on trends and patterns over time with less vulnerability to errors. The government through machine learning approach in asset valuation can identify and segment taxable properties in the housing markets based on real-time geographic data. The potential of expanding the revenue base through rating assessment and enumeration devoid of omission is one of the merit of computerized valuation. It is therefore prudent for the government to collaborate and implement policies that seeks to build the capacity of experts in valuation like the Estate Surveyors and Valuers. Combining the practical knowledge of Estate surveyors and Valuers with the operational efficiency of machine learning in valuation will promote sound management of public assets in Nigeria.
