BanditRLlib
Lean gate passed before this site build; local proof declarations are shown as compiled.Lean-verified build · exact declarations linked.

Lean module · Tsallis-FTRL

BanditRLProof.TsallisOracleRestartScoreAlignment

This module identifies every actual restart probability and stored-reward importance-weighted loss with the corresponding scheduled surface after shifting to the current epoch's local time. It then exposes the pathwise FTRL certificate on any explicit contiguous prefix of one restart epoch.

Module map

Declarations
17
Placeholders
0

Imports

BanditRLProof.TsallisOracleRestartExpectedRegret

Imported by

BanditRLProof, BanditRLProof.TsallisOracleRestartExpectedStability

Declarations

Open an item to read its exact compact statement and source link. Detailed teaching notes are linked when registered.

def BanditRLProof.Tsallis.oracleRestartShiftedTrajectory Compiled

Shift a generated trajectory so that `start` becomes local time zero.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.oracleRestartShiftedTrajectory

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

def oracleRestartShiftedTrajectory {Env : Type u} {Action : Type v} (start : Nat) (sample : Env × ((k : Nat) -> Action × Real)) : Env × ((k : Nat) -> Action × Real)
theorem BanditRLProof.Tsallis.oracleRestartShiftedTrajectory_fst Compiled

No declaration docstring is present; use the chapter context and exact statement below.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.oracleRestartShiftedTrajectory_fst

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem oracleRestartShiftedTrajectory_fst {Env : Type u} {Action : Type v} (start : Nat) (sample : Env × ((k : Nat) -> Action × Real)) : (oracleRestartShiftedTrajectory start sample).1 = sample.1
theorem BanditRLProof.Tsallis.oracleRestartShiftedTrajectory_snd_apply Compiled

No declaration docstring is present; use the chapter context and exact statement below.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.oracleRestartShiftedTrajectory_snd_apply

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem oracleRestartShiftedTrajectory_snd_apply {Env : Type u} {Action : Type v} (start : Nat) (sample : Env × ((k : Nat) -> Action × Real)) (localTime : Nat) : (oracleRestartShiftedTrajectory start sample).2 localTime = sample.2 (start + localTime)
theorem BanditRLProof.Tsallis.OracleRestartSchedule.monotone_start Compiled

Restart epoch starts are monotone in actual time.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.OracleRestartSchedule.monotone_start

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem OracleRestartSchedule.monotone_start (schedule : OracleRestartSchedule) : Monotone schedule.start
theorem BanditRLProof.Tsallis.OracleRestartSchedule.start_start Compiled

Every epoch identifier visited by the schedule is a fixed point of `schedule.start`.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.OracleRestartSchedule.start_start

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem OracleRestartSchedule.start_start (schedule : OracleRestartSchedule) (t : Nat) : schedule.start (schedule.start t) = schedule.start t
theorem BanditRLProof.Tsallis.OracleRestartSchedule.start_eq_of_between Compiled

A restart schedule cannot leave an epoch and later return to it.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.OracleRestartSchedule.start_eq_of_between

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem OracleRestartSchedule.start_eq_of_between (schedule : OracleRestartSchedule) {epoch localTime t : Nat} (ht : schedule.start t = epoch) (hepoch : epoch <= localTime) (hlocalTime : localTime <= t) : schedule.start localTime = epoch
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisProbabilityAtTime_eq_scheduled_shift Compiled

At every actual time, the restarted probability is exactly the scheduled probability at local time `t - schedule.start t` on the shifted trajectory.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisProbabilityAtTime_eq_scheduled_shift

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisProbabilityAtTime_eq_scheduled_shift {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (t : Nat) (sample : Env × ((k : Nat) -> Action × Real)) : sampledOracleRestartHalfTsallisProbabilityAtTime arms harms eta schedule t sample = sampledScheduledHalfTsallisProbabilityAtTime arms harms eta (t - schedule.start t) (oracleRestartShiftedTrajectory (schedule.start t) sample)
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedEstimatedLossAt_eq_scheduled_shift Compiled

The stored-reward restart estimator is the scheduled stored-reward estimator at the current epoch's local time.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedEstimatedLossAt_eq_scheduled_shift

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisObservedEstimatedLossAt_eq_scheduled_shift {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (t : Nat) (sample : Env × ((k : Nat) -> Action × Real)) : sampledOracleRestartHalfTsallisObservedEstimatedLossAt arms harms eta schedule t sample = sampledScheduledHalfTsallisObservedEstimatedLossAt arms harms eta (t - schedule.start t) (oracleRestartShiftedTrajectory (schedule.start t) sample)
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisProbabilityAtTime_add_eq_scheduled_of_start_eq Compiled

On a fixed epoch fiber, the restarted probability is the scheduled local probability on the trajectory shifted by that epoch.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisProbabilityAtTime_add_eq_scheduled_of_start_eq

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisProbabilityAtTime_add_eq_scheduled_of_start_eq {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (epoch localTime : Nat) (sample : Env × ((k : Nat) -> Action × Real)) (hstart : schedule.start (epoch + localTime) = epoch) : sampledOracleRestartHalfTsallisProbabilityAtTime arms harms eta schedule (epoch + localTime) sample = sampledScheduledHalfTsallisProbabilityAtTime arms harms eta localTime (oracleRestartShiftedTrajectory epoch sample)
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedEstimatedLossAt_add_eq_scheduled_of_start_eq Compiled

On a fixed epoch fiber, the stored-reward restarted estimator is the scheduled local estimator on the shifted trajectory.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedEstimatedLossAt_add_eq_scheduled_of_start_eq

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisObservedEstimatedLossAt_add_eq_scheduled_of_start_eq {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (epoch localTime : Nat) (sample : Env × ((k : Nat) -> Action × Real)) (hstart : schedule.start (epoch + localTime) = epoch) : sampledOracleRestartHalfTsallisObservedEstimatedLossAt arms harms eta schedule (epoch + localTime) sample = sampledScheduledHalfTsallisObservedEstimatedLossAt arms harms eta localTime (oracleRestartShiftedTrajectory epoch sample)
def BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret Compiled

Stored-reward estimated regret on an explicit inclusive prefix of one restart epoch, indexed by local times `0, ..., localHorizon`.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

noncomputable def sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (q : Action -> Real) (epoch localHorizon : Nat) (sample : Env × ((k : Nat) -> Action × Real)) : Real
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret_eq_scheduled Compiled

A contiguous restart-epoch prefix is definitionally the existing scheduled estimated regret on the shifted trajectory.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret_eq_scheduled

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret_eq_scheduled {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (q : Action -> Real) (epoch localHorizon : Nat) (sample : Env × ((k : Nat) -> Action × Real)) (hstart : forall localTime, localTime <= localHorizon -> schedule.start (epoch + localTime) = epoch) : sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret arms harms eta schedule q epoch localHorizon sample = sampledScheduledHalfTsallisEstimatedRegret arms harms eta q localHorizon (oracleRestartShiftedTrajectory epoch sample)
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret_pointMass_le_stability_add_penalty Compiled

Pathwise FTRL certificate for any explicit contiguous prefix of one restart epoch.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret_pointMass_le_stability_add_penalty

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret_pointMass_le_stability_add_penalty {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (epoch localHorizon : Nat) (sample : Env × ((k : Nat) -> Action × Real)) {best : Action} (hbest : best ∈ arms) (hstart : forall localTime, localTime <= localHorizon -> schedule.start (epoch + localTime) = epoch) (heta : forall localTime, localTime <= localHorizon -> 0 < eta localTime) (hetaMono : forall localTime, localTime < localHorizon -> eta (localTime + 1) <= eta localTime) : sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret arms harms eta schedule (pointMass best) epoch localHorizon sample <= (Finset.range (localHorizon + 1)).sum (fun localTime => sampledScheduledHalfTsallisPotentialStabilityAtTime arms harms eta (oracleRestartShiftedTrajectory epoch sample) localTime) + halfTsallisPotentialMass arms (sampledScheduledHalfTsallisProbabilityAtTime arms harms eta 0 (oracleRestartShiftedTrajectory epoch sample)) / eta localHorizon - 1 / eta localHorizon
theorem BanditRLProof.Tsallis.exists_oracleRestartEpochRounds_eq_image_range Compiled

Every visited restart epoch fiber is a nonempty contiguous range starting at its epoch identifier.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.exists_oracleRestartEpochRounds_eq_image_range

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem exists_oracleRestartEpochRounds_eq_image_range (schedule : OracleRestartSchedule) (horizon epoch : Nat) (hepoch : epoch ∈ oracleRestartScheduleEpochs schedule horizon) : ∃ localHorizon, oracleRestartEpochRounds schedule.start horizon epoch = (Finset.range (localHorizon + 1)).image (fun localTime => epoch + localTime)
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_eq_localPrefix_of_epochRounds_eq Compiled

An explicit range representation of an actual epoch fiber identifies its stored-reward regret with the local-prefix surface.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_eq_localPrefix_of_epochRounds_eq

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_eq_localPrefix_of_epochRounds_eq {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (q : Action -> Real) (horizon epoch localHorizon : Nat) (sample : Env × ((k : Nat) -> Action × Real)) (hRounds : oracleRestartEpochRounds schedule.start horizon epoch = (Finset.range (localHorizon + 1)).image (fun localTime => epoch + localTime)) : sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret arms harms eta schedule q horizon epoch sample = sampledOracleRestartHalfTsallisLocalPrefixObservedEstimatedRegret arms harms eta schedule q epoch localHorizon sample
theorem BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_pointMass_le_stability_add_penalty_of_epochRounds_eq Compiled

Pathwise FTRL certificate for an actual epoch once its finite fiber is presented as a contiguous local-time range.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_pointMass_le_stability_add_penalty_of_epochRounds_eq

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_pointMass_le_stability_add_penalty_of_epochRounds_eq {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (horizon epoch localHorizon : Nat) (sample : Env × ((k : Nat) -> Action × Real)) {best : Action} (hbest : best ∈ arms) (hRounds : oracleRestartEpochRounds schedule.start horizon epoch = (Finset.range (localHorizon + 1)).image (fun localTime => epoch + localTime)) (heta : forall localTime, localTime <= localHorizon -> 0 < eta localTime) (hetaMono : forall localTime, localTime < localHorizon -> eta (localTime + 1) <= eta localTime) : sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret arms harms eta schedule (pointMass best) horizon epoch sample <= (Finset.range (localHorizon + 1)).sum (fun localTime => sampledScheduledHalfTsallisPotentialStabilityAtTime arms harms eta (oracleRestartShiftedTrajectory epoch sample) localTime) + halfTsallisPotentialMass arms (sampledScheduledHalfTsallisProbabilityAtTime arms harms eta 0 (oracleRestartShiftedTrajectory epoch sample)) / eta localHorizon - 1 / eta localHorizon
theorem BanditRLProof.Tsallis.exists_sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_pointMass_le_stability_add_penalty Compiled

Every visited epoch admits a local horizon whose cardinality is the actual fiber cardinality and whose stored-reward regret satisfies the pathwise FTRL stability-plus-penalty certificate.

Used in these reading views: Bandit Book · Online Learning Book

8. Tsallis-FTRL, corruption, and nonstationarity

Canonical node identitydeclaration:BanditRLProof.Tsallis.exists_sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_pointMass_le_stability_add_penalty

Reading membership is not a proof dependency. Exact assumptions remain in the Lean statement.

theorem exists_sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret_pointMass_le_stability_add_penalty {Env : Type u} {Action : Type v} [DecidableEq Action] (arms : Finset Action) (harms : arms.Nonempty) (eta : Nat -> Real) (schedule : OracleRestartSchedule) (horizon epoch : Nat) (hepoch : epoch ∈ oracleRestartScheduleEpochs schedule horizon) (sample : Env × ((k : Nat) -> Action × Real)) {best : Action} (hbest : best ∈ arms) (heta : forall localTime, 0 < eta localTime) (hetaMono : forall localTime, eta (localTime + 1) <= eta localTime) : ∃ localHorizon, oracleRestartEpochRounds schedule.start horizon epoch = (Finset.range (localHorizon + 1)).image (fun localTime => epoch + localTime) ∧ (oracleRestartEpochRounds schedule.start horizon epoch).card = localHorizon + 1 ∧ sampledOracleRestartHalfTsallisObservedScheduleEpochEstimatedRegret arms harms eta schedule (pointMass best) horizon epoch sample <= (Finset.range (localHorizon + 1)).sum (fun localTime => sampledScheduledHalfTsallisPotentialStabilityAtTime arms harms eta (oracleRestartShiftedTrajectory epoch sample) localTime) + halfTsallisPotentialMass arms (sampledScheduledHalfTsallisProbabilityAtTime arms harms eta 0 (oracleRestartShiftedTrajectory epoch sample)) / eta localHorizon - 1 / eta localHorizon