C4cast.com, Inc.

Lotus Development Corporation

Stephen A Klein

William Palmer Jennings

G Michael Phillips

M Chapman Findlay, Iii

Mark E Rice

David A Rolfe

William T Gross

Lawrence S Gross

Gordon Michael Phillips

The value of a target variable is predicted by obtaining historical values for the target variable at each of several time points and obtaining previously predicted values and currently predicted values for each of several predictor variables, the predictor variables being different from the target variable. Values are assigned to parameters of a forecasting model to obtain the best fit of the previously predicted values for the predictor variables to the historical values for the target variable. Finally, a value of the target variable is predicted from the currently predicted values for at least a subset of the predictor variables using the forecasting model and the values assigned to the parameters of the forecasting model.

Community-selected content

Provided are, among other things, systems, methods and techniques for forecasting information. One representative embodiment involves obtaining data that has been provided by each of a number of individual people; dividing the people into clusters by utilizing a statistical clustering technique; calculating statistics of the data in each of at least some of the clusters; and combining the statistics for such clusters so as to obtain a combination forecast.

Analysis and prediction of data using clusterization

Provided is a technique for evaluating an asset, such as a share of stock in a corporation, by processing historical data for value of an asset and historical data values for various exogenous variables to obtain a formula for calculating a measure of a tendency of the value of the asset to change as a result of changes in the data values for the exogenous variables, the formula being a function of the exogenous variables. Projected data values are then obtained for the exogenous variables, and a measure of the tendency of the value of the asset to change based on a change in at least one of the exogenous variables is estimated using the formula and the input projected data values.

Sensitivity/elasticity-based asset evaluation and screening

Provided are, among other things, systems, methods and techniques for providing resources to participants over an electronic network. In one representative embodiment, a collection of resources is maintained, such that both the collection and the resources can be accessed by a participant over the electronic network at any given time; points are assigned to individual resources based on an amount of participant access of the individual resources over the electronic network; and the collection is modified based on the points assigned to the resources.

Provided is a technique forecasting the direction in which the price of an asset will move by identifying a group of exogenous variables that are likely to influence observed prices of an asset. Then, historical data for values of the exogenous variables and historical data for the observed prices of the asset over a time period are processed to obtain a formula for calculating price estimates for the asset as a function of the exogenous variables. The formula is calculated using an input set of observed values for the exogenous variables at given point in time, so as to obtain a price estimate for the asset at the given point in time. Then, a similarity measure is determined by comparing the price estimate for the asset at the given point in time to the observed price for the asset at the given point in time. Finally, a direction in which the observed price of the asset will move is forecast based on the similarity measure.

Asset price forecasting

Provided is a technique for displaying information. Initially, a number of data points is obtained, each data point including an estimated statistic. A measure of the statistical significance for each estimated statistic also is obtained, and a graph of the data points is displayed. According to this aspect of the invention, a display characteristic (such as intensity, size, brightness, hue, saturation or any other color characteristic) of each data point is a function of the measure of statistical significance of the estimated statistic included in such data point.

Significance-based display

Forecasting using interpolation modeling

Provided is a technique for classifying stocks into business sectors by calculating, for each of multiple exogenous variables, a measure of a tendency for a value of a stock to change as a result of a change in a data value for each such exogenous variable. The foregoing step is then repeated for each of several different stocks. Finally, the different stocks are grouped into different sectors based on similarities of such measures of tendency to change.

Identifying industry sectors using statistical clusterization

Combination forecasts are generated using predictions obtained from a group of forecasters. The forecasters are first divided into a number of pre-defined clusters, which typically will have been formed using statistical clustering techniques. In particular, clusters of forecasters can be formed based on similarities of the forecasters' predictions. Then, statistical data are calculated for each pre-defined cluster (e.g., measures of central tendency and dispersion). Finally, the statistical data for all the pre-defined clusters are combined so as to obtain a combination forecast.

Combination forecasting using clusterization

Resources are provided to participants over an electronic network by maintaining a collection of resources that can be accessed by a participant over the electronic network at a given time. Points are assigned to each resource based on participant access of the resource and the collection is modified based on the points assigned to each resource. Alternatively, or in addition, participants may be permitted to rate the resources, with points assigned to each resource based on participant rating of the resource. The collection of resources is then modified based on assigned points for each resource.

Provision of informational resources over an electronic network

Provided is a contest that produces forecasting data for predesignated variables whose values change over time. Initially, participant registrations are accepted, and the participants are permitted to submit predictions of values, projected at plural different time points, for at least one of several predesignated variables. Then, the participants receive an overall ranking based on their relative accuracies (e.g., percentile rankings) in individual prediction events. Additional features of the invention include: requiring demographic information as a condition to registration, rewarding participants for updating their predictions as early as possible, basing overall ranking on both accuracy and accuracy consistency in individual prediction events, basing the overall ranking on how soon the participants' final predictions were made before certain closing time points, and permitting participants to submit estimates of their own prediction uncertainty.

Forecasting contest

Initially, a graph is electronically displayed, the graph including a historical portion that includes historical values of the variable over time and also including a future portion. Then, a participant is permitted to designate a point on the future portion of the graph (e.g., by using an input device such as a mouse, a touch-sensitive display screen or the like) and the designated point is converted into a predicted value for the variable at a realization time.

Prediction input

The present invention provides apparatus and methods for determining the boundaries of arbitrarily shaped regions on a computer display system having a central processing unit (CPU). The CPU is coupled to a display having a plurality of selectively enabled and disabled display elements arranged in a matrix, whereby each display element is identified by a unique X,Y address. A memory coupled to the CPU includes a plurality of memory cells, wherein the cells correspond to a display element on the display. An initial X,Y address is selected on the display by a user employing a cursor control to identify an area on the display where the boundaries of a region are to be determined. The initial X,Y address selected by the user is defined as a seed cell. The CPU determines if the initial seed cell corresponds to a disabled memory cell and blank (disabled) display element, and in such event increments the initial X,Y address to search outwardly a predetermined maximum distance to adjacent memory cells until an enabled (non-blank) cell is located. This enabled cell is set as a new seed cell. The CPU then searches radially outward in M directions from the new seed cells X,Y address for N consecutive disabled (blank) memory cells and sets in each of these directions the last non-blank (enabled) memory cell prior to the N consecutive disabled cells, as boundary cells. An initial rectangular region is defined through at least two of these boundary cells, and the CPU determines if P consecutive rows and columns of blank (disabled) memory cells bound the initial rectangular region. If not, the CPU selectively extends the boundaries of the initial rectangular region in X and Y directions until the region is bounded by P consecutive rows and columns of disabled cells. Accordingly, the boundaries of a region on a display are determined.

Growing community of inventors

Pasadena, CA, United States of America

Stephen A Klein