Next: About this document
Up: Two-sided learning in an
Previous: References
-
: the cost of delivering
articles to consumer i.
- F: subscription fee.
- f: normalized subscription fee = F/N.
-
: Any nonlinear function
for
.
- G(q;w): cumulative distribution function of g(q;w).
- g(q;w): the probability density function from which the consumer
valuations w are drawn. q is the distribution
parameter which may vary from one consumer to another.
- h(r;q): the probability distribution for the consumers' q parameter.
r is the distribution parameter.
- i: index of consumers as in consumer i.
- j: index of valuations as in
.
- K: set of articles.
- k: index of articles;
.
- M: total number of consumers in the system.
- m: proportion of subscribing consumers.
- N: number of articles generated by the producer in each subscription
period.
- n: dimension of problem space as in
.
-
: the price schedule =
.
- P(k): the price charged for a subset of
of
the N articles.
-
: the vector of probabilities that any k is the expected
surplus maximizing number of articles.
- q: the distribution parameter in g(q;w).
- r: the distribution parameter in h(r;q).
- s: consumer surplus.
- t: index of the subscription period.
-
: consumer valuation of article j.
- x:
.
-
: reflection coefficient in the amoeba algorithm.
-
: expansion coefficient in the amoeba algorithm.
-
: normalized production cost assuming linear production cost;
.
-
: normalized price-per-item assuming linear price schedule;
.
-
: contraction coefficient in the amoeba algorithm.
-
: simplex shrink/expand coefficient in the
amoeba algorithm.
-
: hard threshold function. Equals 1 if consumer i
subscribes, else it is 0.
-
: parameter for the exponential distribution of consumer valuations;
.
-
: normalized expected profit;
.
-
: expected profit.
-
: the mean of the N valuations received
during the subscription period t.
-
: the consumer's ``flightiness'' factor.
Next: About this document
Up: Two-sided learning in an
Previous: References
kephart
Thu Nov 18 11:46:57 EST 1999