Rates at which the market will "absorb" a product, the rate at which units will be leased or sold.
A loss in value due to physical deterioration, functional obsolescence and economic or locational disadvantages.
Ad Val Orem
A Latin term used to describe the taxing of property according to value.
The amount paid for property, plus all subsequent capital expenditures made to improve it, minus all tax deductions for depreciation of cost recovery allowances.
Adjusted internal rate of return (AIRR)
An internal rate of return analysis in which different reinvestment and discount rates for both positive and negative cash flows have been specified. The adjusted internal rate applies a "safe rate" to all negative cash flows, discounts them to time period zero, and adds them to the initial investment. The safe rate is the rate that could be earned on the funds until needed to cover negative cash flows. A market rate is applied to all positive cash flows, which are carried forward to the end of the investment holding period. The market rate is the rate that can be earned by investing the positive cash flows in other investments. The AIRR equals the internal rate of return (IRR) that equates the present value figure to the future value figure. The AIRR is usually less than the IRR for the same property because funds are assumed to be reinvested at a lower rate. Also called the modified internal rate of return. See also financial management rate of return, internal rate of return.
Adjusted rate of return
Modified version of the internal rate of return, designed to eliminate problems associated with negative cash flows.
Any series of periodic payments received or paid at regular intervals.
Artificial neural network (ANN)
A model made to simulate a biological neural system; it may model brain processes or brain capabilities.
Artificial intelligence (AI)
A branch of computer science in which computers are used to model or simulate some portion of human reasoning or brain activity. The behavior of AI systems is based on sequential processing.
A training algorithm for multilayer perceptrons. Reliable and well-known, although significantly slower than some of the more modern algorithms. A supervised learning method in which an error signal is fed back through the network altering weights as it goes, in order to prevent the same error from recurring.
Band of investment technique
A technique used to calculate the weighted average cost of capital for a property. The cost of each source of funds is weighted by a factor equal to the proportion of total funds that will be derived from that source.
The sum of the weighted average cost of capital and the equity buildup factor used in order to determine the appropriate overall capitalization rate.
Networks based on Bayes' theorem, on the inference of probability distributions from data sets.
A function using the logical operators AND, OR, and NOT.
The relationship between cash expenditure requirements and gross revenue from an investment project. Sometimes called a default ratio.
An expenditure of funds that extends the useful life of a capital asset or adds to its value.
The raising of debt or equity funds for real estate projects.
Products destined to be employed in the production of other goods or services.
The ratio between net income from a real estate investment and the value of the investment. Cap Rate = NOI/Value.
To add an amount to the tax basis of a property.
The portion of an overall capitalization rate comprised of recovery of the owner's invested capital.
Cash-on-cash rate of return
The first year's expected after-tax cash flow, divided by the initial cash outlay used to acquire the property.
The simplest kind of symbolic concept. Columns, or modules, may be the physical implementations of clusters.
Coefficient of correlation
A measure of the extent to which the values of variables in a sample or a population are interdependent.
Coefficient of determination
A measure of the percent of variation in the dependent variable associated with variation in the value of the interdependent variable. In regression analysis, designated by the term r-square (r2).
Coefficient of multiple correlation
A measure of the percentage of variance in the dependent variable that is explained by variation in the interdependent variables.
Coefficient of serial correlation
A measure of the degree to which outcomes in subsequent periods are related time periods. Possible coefficients range from zero to plus-or-minus one.
Coefficient of variation
Standard deviation of the distribution of possible outcomes, divided by the expected outcome.
The summation of principal and compounded interest over a specified holding period.
Interest income attributable to previously accrued interest that has been left on deposit, or interest paid on interest.
Conjugate gradient descent
A fast training algorithm for multilayer perceptrons which proceeds by a series of line searches through error space. Succeeding search directions are selected to be conjugate (non-interfering).
A theory of learning which states that two signals converging on a neuron at about the same place and time will cause highly increased membrane sensitivity, increasing the strength of the connection between the neurons in a nonlinear manner.
Something of value given in exchange for a promise.
Consumer price index
An index of changes in the price of a representative "market basket" of consumer goods, relating the current price to that in a designated base year.
The changing state of the network as it moves toward a stable state after an input pattern has been applied.
A measure of the tendency of two variables to vary in such a way as to preserve some relationship between them.
The process of gradually extinguishing a debt by a series of periodic payments to the creditor.
The percentage of the original principal amount that must be paid annually in order to fully repay interest and principal over the term of the loan. The constant can be expressed as an annual percentage or monthly percentage. Sometimes called a debt service constant.
Debt coverage ratio
The relationship between a project's annual net operating income and the obligation to make principal and interest payments on borrowed funds. Debt coverage ratios are often employed to reduce lender's risk regarding mortgage loans.
Payments to a lender. Debt service obligations may involve payment of interest only or both principal and interest so as to fully or partially amortize a debt over a specified term.
An economic term that refers to the entire range of relationships between price and supply/quantity.
A graphic illustration of the relationship between price and the quantity of a good or service that buyers will consider.
A table that relates quantity demanded to a good's or service's price, at all relevant prices.
Decline in an asset's value or useful life, due to wear, tear, action of the elements or obsolescence.
A quantity problem-solving approach in which certainly of an answer exists. Functions of networks can be deterministic or stochastic. Deterministic networks use variables which are bounded by limits in their functions. Stochastic networks use random variables in some cases.
In radial units, a figure multiplied by the radial exemplar's squared distance from the input pattern to generate the unit's activation level, before submission to the activation function.
Discounted cash-flow approach
An investment evaluation technique that incorporates adjustments for both volume and timing of anticipated future cash flows and is generally accepted as the most desirable approach for evaluating opportunities.
Expressing anticipated future cash flows as present-value equivalents.
A reduction in net loan proceeds to make the effective interest rate equal the lender's current yield requirement.
A variable that has a distinct identity or value, in contrast to a continuous valued variable. A discrete state neuron can have only 1 value from a certain set of values. Discrete time describes the nature of the variable used for time, t, where t is an integer and time is counted in steps t+1, t+2, etc. With each step, the value for the neuron is updated.
Loss in value due to inappropriate location.
Markets in which all relevant information is immediately and fully reflected in market prices. Participants in efficient markets are unable to consistently achieve above-average market yields. The hypothesis that a market is completely efficient is referred to as the strong form of the efficiency hypothesis.
During iterative training, a single pass through the entire training set, followed by testing of the verification set.
The price at which there will be sufficient quantity of a product to satisfy the desires of all consumers at that price but with no surplus remaining on the market; the market clearing price.
Estimate of the likelihood of achieving explicit project objectives through a proposed course of action, given a specific set of constraints and limited resources.
A network in which neurons can take their inputs from any other neuron, including their own output. Since a neuron is allowed to connect to any other neuron, feedback networks usually have only one layer. To compute a result, a feedback network must repeatedly compute the result until the neurons settle into a stable state. There is no way to predict how long this will take.
Neural networks with a distinct layered structure, with all connections feeding forward from inputs towards outputs. Sometimes used as a synonym for multilayer perceptrons.
Financial management rate of return (FMRR)
A modification of the internal rate of return, designed to eliminate problems encountered when negative cash flows are included in the forecast.
A measure of how well a property performs its intended function.
The loss of functional efficiency due to defective or dated design. This reduces a property's competitive position relative to more functionally efficient structures and may eventually lead to difficulties and abandonment.
An investment analysis technique that emphasizes investigation of the underlying business activity being undertaken by the firm whose securities are being considered.
Processing information that is ambiguous. Fuzzy sets may overlap one another (e.g. something is both sweet and sour). Fuzzy logic uses the operations AND, OR and NOT.
A search algorithm which locates optimal binary strings by processing an initially random population of strings using artificial mutation, crossover and selection operators, in an analogy with the process of natural selection.
Optimization techniques for non-linear functions (e.g., the error function of a neural network as the weights are varied) which attempt to move incrementally to successively lower points in search space, in order to locate a minimum. It is the simplest method for finding a minimum. This algorithm works by moving downhill in the steepest direction, much like a downhill skier. Back propagation uses gradient descent to find the solution with the smallest amount of error.
Gross income multiplier
Evaluation technique that describes the relationship between the sales price and gross revenue. Sometimes called a gross rent multiplier.
A layer of neurons in an artificial neural network which does not connect to the outside world but connects to other layers of neurons. All layers of a neural network except the input and output layers, which provide its non-linear modeling capabilities.
A neuron in a hidden layer.
In regression or correlation analysis, a variable whose value is thought to be determined by factors other than those under analysis, but which is thought to affect the value of one or more other variables (the dependent variables) in the analysis.
Internal rate of return (IRR)
A financial analysis technique that involves setting net present value at zero and finding a discount rate to satisfy the equality condition; that is, the discount rate that makes present value exactly equal to required initial cash outlay. This method fails when there are negative cash flows.
Projecting a curve between known data points to infer the value of a function at points between.
A network which gathers a clearer set of data from a noisy set.
Commitment of money or other assets in expectation of financial gain.
Any person or entity who takes an equity position in real estate for use in a trade or business or for production of income.
Adding a small random amount to the weights in a neural network, in an attempt to escape a local optima in error space.
The probability of joint occurrence or two or more events. The probability that both event A and event B will occur equals the probability that A will occur times the probability that B will occur, given that A occurs. This relationship is sometimes referred to as the multiplicative law of probability.
A group of neurons which share a functional feature. Most of the human cortex is made of many layers of neurons. In each layer, the neurons are densely interconnected. A neural network consists of neuron layers which are interconnected. The manner in which the neurons interconnect determines the type of processing that will occur. (See feedback and feed forward networks). Neurons are located in one of three types of places: the input layer, the output layer, or the hidden layers. The input neurons receive data from the outside world. The output neurons send information out to us, or to something else. The hidden neurons are all the neurons in between. We do not see their inputs and outputs because they connect only to other neurons.
A factor to scale all corrections while learning; intended to improve the speed of convergence of the network.
A property owner who transfers certain rights for a limited period to a tenant. Generally referred to as a landlord.
The holder of a leasehold interest in a property. Generally referred to as a tenant.
A non-linear optimization algorithm which uses a combined strategy of linear approximation and gradient-descent to locate a minimum, actively switching between the two according to the success or failure of the linear approximation; a so-called model-trust region approach.
Claim against a property that allows the proceeds from a forced sale of the property to be used to satisfy the debt.
Linear activation function
A small activation function; the unit's output is identical to its activation level.
The early associative memory network model developed by Anderson and Kohonen. It is the first model that used neurons with linear transfer functions instead of threshold logic units. The neurons respond to changes in the inputs by changing the firing rate of their outputs. The network maps similar inputs to similar outputs, leading automatically to the ability to generalize. Simple linear associators use Hebb's Rule, which allows the network to distinguish between patterns which are mutually orthogonal. If not all the input patterns are orthogonal, there is interference among them, and the network does not correctly learn all the patterns. Linear associators use the Delta Rule, which requires that all the input patterns be linearly independent for perfect learning.
Approximation of a discriminant function or regression function using a hyperplane. Can be globally optimized using "simple" techniques, but does not adequately model many real-world problems.
Linear transfer function
A function in which the output is equal to a gain value times the difference between the net input value less the center value. The center value is the value of net input at which the output is equal to zero.
Relationships requiring the movement of goods or people from one location to another.
Ability to convert an asset to cash without incurring loss.
Marginal cost of capital
The cost of an additional dollar of new capital funds.
Marginal cost of production
The cost of adding one or more unit per period to one's rate of production.
Market demand curve
Curve showing the amount of an economic good or service that will be demanded at various price levels. Demand curves are typically downward sloping, indicating that as price increases, demand increases.
The rent a property would command on the open market if it were currently vacant and available.
In statistics, the arithmetic average. In Real Estate, how the landlord is when the tenant fails to pay rent.
Theory of small economic units. Generally referred to as the theory of the firm.
Modified internal rate of return (MIRR)
A variant of the internal rate of return, intended to eliminate the multiple root problem by discounting all negative cash flows back to the time an investment commitment must be made and by compounding all positive cash flows forward to the end of the final year of the investment holding period. The modified internal rate of return is the discount rate that equates the present value of all negative cash flows with the future value of all positive cash flows.
Modified internal rate of return
See adjusted internal rate of return.
Use of control over the money supply to stimulate or dampen economic growth.
Feedforward neural networks having linear PSP functions and (usually) non-linear activation functions.
Statistical technique used to measure the association between a dependent variable and multiple independent variables.
Adam Smith's concept of long-run, market-determined price. Smith held that the price of all goods and services will, over the long run, equal the cost of production.
The physical vacancy rate at which there is no upward or downward pressure on rents (vacancy equilibrium point).
Net income multiplier
(See also gross income multiplier and income multiplier analysis.) Property market value expressed as a multiple of its net operating income.
Net operating income (NOI)
Effective gross revenue less operating expenses.
Net present value (NPV)
The discounted value of all future cash flows minus the initial cash outlay. A net present value greater than or equal to zero is acceptable.
Highly parallel dynamic system that carries out information processing by means of its overall state response to continuous or initial input. A mathematical model of the brain's neurons. Sometimes thought of as feedback amplifiers connected in parallel. Also referred to as artificial neural system, natural intelligence, and neurocomputer.
A nerve cell in a biological nervous system; a processing element in a neural network. It has a number of inputs and a single output.
Irrelevant or imprecise data present in input patterns; random values added to all the weights to prevent the network from getting stuck in a local energy minima; imprecise information such as seeing most of a box that is partially hidden; imprecise data purposely put in the initial state to improve a network's accuracy.
In neural networks, nonlinearity can take several forms/feedback into the same layer, multi-layer feedback, normalization and competition, or multi-layer feed-forward networks which use non-linear transfer functions.
Cash expenditures required to maintain property in sufficient condition to generate effective gross revenue.
Logical problems arising from real world situations which involve picking the best answer out of many possible ways to solve the problem. Optimization problems are found in engineering, commerce, and perception. Also called explosive or combinatorial problems.
Data points which do not appear to follow the characteristic distribution of the rest of the data. These may reflect genuine properties of the underlying phenomenon (variable), or be due to measurement errors or other anomalies which should not be modeled.
When an iterative training algorithm is run, overfitting which occurs when the algorithm is run for too long or the network is too complex for the problem or the available quantity of data.
The ability to recognize a set of input data instantaneously and without conscious thought, such as recognizing a face. The ability of a neural network to identify a set of previously learned data, even in the presence of noise and distortion in the input pattern.
In statistics, the entire universe of data from which samples are drawn.
Potential gross income
The maximum amount of revenue a property would produce if fully rented at market rates.
Potential gross rent
The amount of rental revenue a property would generate if there were no vacancies.
Present value (PV)
The current value of a payment or series of future payments found by discounting the expected payments by a desired rate of return in order to compensate for the time value of money. See also internal rate of return and net present value.
Data gathered by researchers specifically for the problem on which they are currently working.
Specification of the odds, likelihood of an event, or evidence supporting a conclusion. A mathematical basis for prediction.
Probability density function
A function that gives the likelihood that a random variable has values in the set. Weight vectors get stuck in isolated regions in Kohonen and counter propagation networks. This can be prevented by adding noise to the data, which makes the probability density function positive everywhere. This works, but it is slower than convex combination.
Probabilistic neural networks (PNN)
A type of neural network using kernel-based approximation to form an estimate of the probability density functions of classes in a classification problem. One of the so-called Bayesian networks.
Radial basis functions
A type of neural network employing a hidden layer of radial units and an output layer of linear units, and characterized by reasonably fast training and reasonably compact networks.
The measurement of the mean expectation of one stochastically random dependent variable against another; the tendency for the expected value of one of two jointly correlated random variables to approach more closely the mean value of its set than the other.
Statistical technique used to measure the association among two or more variables.
Agreements between landlord and tenant that reduce actual rental payments below those specified in a lease. A landlord might, for example, give one month's free occupancy, thereby reducing the effective rental rate over the entire occupancy period. Also called concessions, or rental incentives.
Root mean square (RMS) error
Yields a single number which summarizes the overall error. It is the square root of the average individual mean error squared.
A measure of the extent to which casual factors influence outcomes over two or more time periods.
A transfer function which has a high and a low saturation limit, and a proportionality range between. The sigmoid function has an S shape when it is plotted.
Simple linear associator
A linear associator with the simplest version of Hebb's rule. The system is tested by presenting an input pattern without a teaching input and seeing how close the pattern generated on the output layer matches the original teaching input. One of the predictions of the simple linear associator is interference between nonorthogonal patterns.
A representation of a system that imitates the behavior of the system.
A measure of dispersion about the mean of a probability distribution, frequently employed as an indication of risk associated with an investment venture. The square root of the variance.
Standard error of the forecast
The degree of confidence to be placed in a forecast value for a dependent variable. Conceptually similar to the same measure as calculated in simple linear regression.
A function in which the values depend on a probability or random factor.
A function which depends on a random variable. Functions of networks can be deterministic or stochastic. Deterministic networks use variables which are bounded by limits in their functions. Stochastic networks use random noise generators.
The area of electrochemical contact between two neurons. Synapses can be excitatory (causing an increase in the receiving neuron's activation level) or inhibitory (causing a decrease in the activation level).
The relationship between price and the quantity of a product suppliers place on the market during a specified time period, for all possible prices.
Supply and demand
An appraisal principle that states that the value of a property depends on the quantity and price of the property type available in the market and on the number of market participants and the price that they are willing to pay.
Fundamental mathematical rules from which various mathematical operations are derived.
Problems (usually regression) where the objective is to predict later values of a variable or variables from earlier values.
The price at which a transaction actually occurred; the outcome of a bargaining process between buyer and seller.
A measure of dispersion of possible values about the midpoint of a probability distribution of possible outcomes, frequently employed as a measure of risk. The square of the standard deviation.